REVIEW ARTICLE
Epidural analgesia in patients with traumatic rib fractures: a systematic review of randomised controlled trials P. Duch1 and M. H. Møller2 1 2
Department of Anaesthesiology and Intensive Care Medicine, Copenhagen University Hospital Hvidovre, Hvidovre, Denmark Department of Intensive Care 4131, Copenhagen University Hospital Rigshospitalet, Copenhagen, Denmark
Correspondence P. Duch, Department of Anaesthesiology and Intensive Care Medicine, Copenhagen University Hospital Hvidovre, Kettegård Alle 39, Hvidovre 2650, Denmark E-mail: [email protected] Conflicts of interest The authors have no conflicts of interest. Funding This research received no grant from any funding agency in the public, commercial or not-for-profit sectors. Submitted 26 December 2014; accepted 29 December 2014; submission 5 November 2014. Citation Duch P, Møller MH. Epidural analgesia compared with other analgesic interventions in patients with traumatic rib fractures. Acta Anaesthesiologica Scandinavica 2015 doi: 10.1111/aas.12475
Background: Traumatic rib fractures are a common condition associated with considerable morbidity and mortality. Observational studies have suggested improved outcome in patients receiving continuous epidural analgesia (CEA). The aim of the present systematic review of randomised controlled trials (RCTs) was to assess the benefit and harm of CEA compared with other analgesic interventions in patients with traumatic rib fractures. Methods: We performed a systematic review with meta-analysis and trial sequential analysis (TSA). Eligible trials were RCTs comparing CEA with other analgesic interventions in patients with traumatic rib fractures. Cumulative relative risks (RRs) and mean differences (MDs) with 95% confidence intervals (CIs) were estimated, and risk of systematic and random errors was assessed. The predefined primary outcome measures were mortality, pneumonia and duration of mechanical ventilation. Results: A total of six trials (n = 223) were included; all were judged as having a high risk of bias. In the conventional metaanalyses, there was no statistically significant difference in mortality (RR 2.18, 95% CI 0.21–22.42; P = 0.51; I2 = 0%), duration of mechanical ventilation (MD −7.53, 95% CI −16.32 to 1.26; P = 0.09; I2 = 91%) or pneumonia (RR 0.49, 95% CI 0.19–1.25; P = 0.13; I2 = 0%) between CEA and other analgesic interventions. Subgroup analyses and sensitivity analyses, including TSA confirmed the results. Conclusion: The quality and quantity of evidence for the use of CEA in patients with traumatic rib fractures is low, and there is no firm evidence for benefit or harm of CEA compared with other analgesic interventions. Well-powered RCTs with low risk of bias reporting clinically relevant patient-centred outcome measures are needed.
Editorial comment: what this article tells us In this well-performed systematic review with meta-analysis and trial sequential analysis, it was somewhat surprisingly demonstrated that there is no firm evidence for benefit or harm of continuous epidural analgesia compared with other analgesic interventions in patients with traumatic rib fractures.
Acta Anaesthesiologica Scandinavica (2015) © 2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
bs_bs_banner
1
P. DUCH AND M. H. MØLLER
Rib fractures are a common condition following trauma with a reported incidence of up to 10% in trauma in general,1 and up to 39% in blunt chest trauma.2 Traumatic rib fractures are associated with significant morbidity and mortality,1 and mortality rates reaching 10–16% has been reported.2,3 An estimated one third of patients with traumatic rib fractures develop secondary pulmonary complications2,4–6 with an associated mortality rate as high as 65%.7 Pain is recognised as a contributing factor to adverse outcome in traumatic rib fractures due to pain-induced inadequate respiratory efforts leading to atelectasis, difficulties in clearing secretions and an increased risk of developing pneumonia.3,6,8,9 Consequently, adequate analgesia is considered a core intervention in the management of patients with traumatic rib fractures.6,10–12 Based on observational data from the United States and Canada and older small randomised controlled trials (RCTs), it has been suggested that continuous epidural analgesia (CEA) is associated with statistically significantly increased survival compared with other analgesic interventions in patients with traumatic rib fractures.1,2,6,9,11 However, it is of major importance to evaluate the risk of design errors, systematic errors (bias) and random errors before concluding on the superiority of one intervention over the other.13 Existing efforts on this matter have been both quantitatively and qualitatively limited.14 Uncertainty over whether CEA is superior to other analgesic interventions in patients with traumatic rib fractures therefore exists among clinicians. Consequently, there is a need to systematically and rigorously re-appraise the better analgesic intervention in patients with traumatic rib fractures. The objective of the present systematic review of RCTs was to evaluate the benefit and harm of CEA compared with other analgesic interventions using the Cochrane Collaboration methodology, including an assessment of the risk of bias and random errors [trial sequential analysis (TSA)].
tematic Reviews and Meta-Analyses statement.16 The protocol has pre-experimentally been published in the International Prospective Register of Systematic Reviews, no. CRD42014010492. Eligibility criteria RCTs in patients with one or more traumatic rib fractures were considered for inclusion, irrespective of language, blinding, publication date and publication status. CEA by placement of an epidural catheter in the epidural space followed by either a repeated bolus, or an infusion of a local anaesthetic agent and/or opioid, was considered the experimental intervention. The control group included one of the following four analgesic interventions: 1. Intravenous (IV) analgesia; bolus or infusion of an opioid carried out by clinical staff (nursecontrolled analgesia) and/or by patient (patient-controlled analgesia). 2. Thoracic paravertebral block; placement of an epidural catheter in the thoracic paravertebral space followed by (repeated) bolus and infusion of a local anaesthetic agent and/or opioid. 3. Interpleural block; placement of an epidural catheter in the pleural cavity followed by (repeated) bolus and infusion of a local anaesthetic agent and/or opioid. 4. Intercostal block; bolus of a local anaesthetic agent and/or opioid in relevant intercostal spaces. No restrictions regarding patient age, or dose or duration of drug administration, in the intervention or control group were applied. Trials were permitted to have more than one control group. Exclusion criteria were trials in animals, trials in patients with non-traumatic rib fractures, quasirandomised trials, observational studies, crossover trials, trials comparing different treatment regimens of CEA and trials not reporting the patient-centred outcome measures17 outlined in the protocol. Outcome measures
Methods This systematic review is based on the methodology recommended by the Cochrane Collaboration,15 and the review has been prepared according to the PRISMA: Preferred Reporting Items for Sys-
The predefined primary outcome measures were: 1. All-cause mortality at longest follow-up. 2. Duration of invasive mechanical ventilation (intermittent positive pressure ventilation). Acta Anaesthesiologica Scandinavica (2015)
2
© 2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
EPIDURAL ANALGESIA FOR RIB FRACTURES?
3. Hospital acquired (nosocomial) pneumonia at longest follow-up (defined according to the definitions used in the included trials). Predefined secondary outcome measures were: 1. Pain (assessed according to the methods used in the included trials, e.g. visual analogue scale). 2. Length of stay in the intensive care unit (LOS ICU). 3. Length of stay in hospital (LOS hospital). 4. Pruritus. 5. Obstipation. 6. Delirium. 7. Neurological deficits. 8. Deep vein thrombosis. Search strategy We framed the following focused research question: ‘Is CEA in patients with traumatic rib fractures superior to other analgesic interventions?’ A population, intervention, comparator and outcomes-based question and literature search was created18 (Supplementary material). The following databases were searched for literature: MEDLINE, including MeSH (January 1966 to July 2014), EMBASE (January 1980 to July 2014) and the Cochrane Library (July 2014). We also hand-searched the reference lists of included trials and other systematic reviews of CEA in traumatic rib fractures. The electronic literature search was last updated 7 July 2014. Study selection We independently reviewed all titles and abstracts identified in the literature search and excluded trials that were obviously not relevant. The remaining trials were evaluated in full text. Disagreements between reviewers were resolved through discussion. Data extraction We independently extracted information from each included trial using a data extraction form. The extracted information included trial characteristics (year of publication, duration, follow-up, country), characteristics of the trial participants (number, inclusion and exclusion criteria), intervention and
control characteristics, and the outcomes of interest. If data in the identified trials were inadequate and/or unclear, we sought to contact the authors for further details. Correspondingly, one author provided additional information.8 Risk of bias assessment In order to evaluate the risk of systematic errors in the included trials, we independently assessed the risk of bias as advised by the Cochrane Collaboration,15 including the domains of random sequence generation, allocation concealment, blinding, incomplete outcome data, selective outcome reporting, baseline imbalance and bias due to vested financial interest. If one or more domains were judged as being high or unclear, we classified the trial as having a high risk of bias.15 Statistical analyses For each included trial, we calculated relative risk (RR) with 95% confidence intervals (CI) for dichotomous outcome measures and the mean difference (MD) with 95% CI for continuous outcome measures, and we pooled these measures in conventional cumulative meta-analyses. Heterogeneity among trials was quantified with inconsistency factor (I2)19 and the diversity (D2) statistics.20 If the I2 statistic was 0, we reported the results from a fixed-effect model, and if the I2 statistic was > 0, we reported results from a random-effects model (the most conservative estimate). The risk of random errors in the cumulative meta-analyses was assessed by TSA20–22 (details on TSA are given in Appendix S3). A sensitivity analysis with application of continuity correction in trials of zero events was conducted.23 In addition, a post hoc sensitivity analysis with exclusion of the two oldest trials using CEA with opioid only were carried out.11,24 Risk of small trial bias by means of funnel plot was abandoned because less than 10 trials were included.15 Review Manager 5.2 (The Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen, Denmark) was used for the conventional meta-analyses, and for the TSA, we used the TSA program version 0.9 beta.25 Subgroup analyses A number of predefined subgroup analyses were planned (Appendix S2); however, only the sub-
Acta Anaesthesiologica Scandinavica (2015) © 2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
3
P. DUCH AND M. H. MØLLER
group analysis of the four comparators could be carried out and is consequently reported.
matter.8,26 None of the trials provided information on duration of follow-up.
Results
Participants
Figure 1 summarises the results of the search. Six trials were included: five published in English6,8,9,11,24 and one published in Turkish26 (Table 1). The main reasons for exclusion of trials were (1) no relevant patient-centred outcome measures were reported;27–30 (2) the population of interest was chest trauma in general and not specifically traumatic rib fractures;10,31 and (3) the design of the trial32,33 (Table S1).
The included trials enrolled a total of 223 patients. Four trials included patients from the trauma centre,8,9,11,26 and two trials included patients from the intensive care unit (ICU).6,24 Inclusion and exclusion criteria were overall comparable between trials (Table 1).
Characteristics of trials All of the included RCTs were single-centred trials. Three were conducted in the United States,9,11,24 one in Iran,6 one in Turkey26 and one in India.8 The duration of inclusion varied from 17 to 46 months; two trials did not report on this n = 336 of records identified through database (MEDLINE) searching after duplicates removed
Intervention and comparators Three of the six trials evaluated CEA with a local anaesthetic agent and an opioid,6,9,26 whereas one trial evaluated CEA without additional opioid,8 and two trials evaluated CEA with opioid alone.11,24 The comparator was IV analgesia in four trials,9,11,24,26 paravertebral block in one trial8 and intercostal block in one trial.6 None of the included trials used interpleural block as comparator. n = 4 of additional records identified through hand-search of the reference lists and other sources according to our search strategy
n = 340 of records screened on basis of title and abstract
n = 326 of records excluded
n = 14 of full-text articles assessed for eligibility
n = 8 of full-text articles excluded: Not RCT (n = 1)
n = 6 of studies included in qualitative synthesis
n = 6 of studies included in quantitative synthesis (metaanalysis and trial sequential analysis)
Cross-over design (n = 1) Not relevant end point (n = 1) Chest trauma in general (not rib fractures) (n = 2) Inadequately reported end points (n = 3)
Fig. 1. Study flow diagram. RCT, randomised controlled trial. Acta Anaesthesiologica Scandinavica (2015)
4
© 2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
ICU
60
Hascehmzadeh et al. 20116
© 2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
Trauma/single India centre
Trauma/single Turkey centre
Surgical ICU
30
27
28
Mohta et al. 20098
Sahin et al. 199326
Ullman et al. 198924
36
Unknown
Unknown
36
17
46
3 or more unilateral rib Not mentioned fractures or flail segment with significant contusion of the chest wall with impaired ventilation. Injuries had to be associated with significant pain and splinting with normal ventilation.
Traumatic brain injury.
Severe altered mental status. Unstable cardiac status. Liver or kidney disease. Spine injury. Pelvic fracture. Abdominal visceral injuries. Traumatic brain injury.
> 18 years > 3 unilateral fractured ribs.
Flail chest
Pregnancy. History of substance abuse. Chronic use of analgesics. Chronic pain. Spine injury. Painful limp injury.
> 18 years > 3 rib fractures and flail chest or > 2 rib fractures and exploratory laparotomy/ pulmonary contusion. Normal mental status
Meperidin IV, PCA
Paravertebral catheter with continuous bupivacain and IV morphine p.r.n.
Continuous IV fentanyl infusion
Intercostal block every 8 h with bupivacain. IV pethidine p.r.n.
IV opioids: Morphine, hydromorphine and fentanyl. Administered by patient-controlled analgesia (PCA) for alert patients and with nurse assistance (NCA) for patients who could not participate in self-administration
Comparator
Thoracic epidural catheter Continuous IV morphine or IV with loading dose morphine morphine every 3–4 hours and fentanyl and p.r.n. continuous morphine.
Epidural catheter with continuous bupivacain and fentanyl infusion
Thoracal epidural catheter with continuous bupivacaine and IV morphine p.r.n.
Lumbar epidural catheter with continuous fentanyl infusion
Thoracic epidural catheter with bupivacain and morphine every 8 h. And IV pethidine p.r.n.
Cerebral injury. Mechanical ventilation. Systemic infection. Fever. Liver or splenic trauma.
> 18 years > 1 rib fracture GCS > 14
Intervention
Spine injury. Severe traumatic brain Epidural catheter with injury. Severe altered mental continuous bupivacaine, status. Unstable pelvic fracture or morphine, and fentanyl open abdomen that would preclude positioning for epidural placement. Cardiac instability or coagulopathy. Active chest wall infection. Acute thoracic aortic transection. Patients whose chest wall pain was manageable with oral opioids or anti-inflammatory medications.
Exclusion criteria (Population)
> 18 years 3 or more rib fractures
GCS, Glasgow Coma Score; ICU, intensive care unit; LOS, length of stay; p.r.n., pro re nata.
USA
Trauma/single USA centre
Mackersie et al. 32 199111
Iran
Trauma/single USA centre
46
Duration of inclusion Inclusion criteria Country (months) (Population)
Bulger et al. 20049
Number of patients in trial Setting
Table 1 Characteristics of the included trials.
Mechanical ventilation LOS ICU LOS Hospital Pruritus Neurological deficit Duration of follow-up unknown
Mechanical ventilation LOS ICU Duration of follow-up unknown
Mortality Pneumonia LOS ICU LOS Hospital Pain Neurological deficit Duration of follow-up unknown
Pneumonia Mechanical ventilation LOS Hospital Pain Pruritus Duration of follow-up unknown
LOS ICU LOS Hospital Pain Duration of follow-up unknown
Mortality Pneumonia Mechanical ventilation LOS ICU LOS Hospital Pruritus Duration of follow-up unknown
Outcomes
EPIDURAL ANALGESIA FOR RIB FRACTURES?
Acta Anaesthesiologica Scandinavica (2015)
5
P. DUCH AND M. H. MØLLER
analysis showed no significant difference in mortality in patients treated with CEA compared with those treated with another analgesic intervention: fixed-effect RR (95% CI) 2.18 (0.21–22.42); P = 0.51; I2 = 0% (Fig. S1). Application of an empirical continuity correction in the no-event trial did not change the result. The subgroup analysis of trials using IV analgesia as comparator versus trials using paravertebral block as comparator could not be conducted because of the no-event trial. TSA could not be conducted due to too few data [diversity (D2) adjusted required information size (DIS) < 5%].
Fig. 2. Risk of bias summary.
Risk of bias No trials were judged to be of low risk of bias in all six domains (Fig. 2). The main reasons for high risk of bias were lack of blinding of patients and personnel, and lack of blinding of outcome assessment. Four trials had high risk of reporting bias, because they did not include results for the key outcome measure mortality.6,11,24,26 One trial had inadequate random sequence generation and allocation concealment,24 and one trial had high risk of other bias because of crossover between intervention and control group.9
Outcome measures All-cause mortality Mortality data were obtained from two trials including 76 patients.8,9 The conventional meta-
Duration of mechanical ventilation Four trials (n = 133) reported data on duration of mechanical ventilation.9,11,24,26 The conventional meta-analysis showed no statistically significant difference in duration of mechanical ventilation in patients treated with CEA compared with patients treated with another analgesic intervention: random-effects MD (95% CI) −7.53 (−16.32 to 1.26); P = 0.09; I2 = 91% (Fig. 3). In the post hoc sensitivity analysis (exclusion of Mackersie et al. and Ullman et al.11,24), the point-estimate was: fixed-effect MD (95% CI) −4.17 (−5.45 to −2.88); P = 0.61; I2 = 0%. The subgroup analysis of the different comparators could not be conducted because all four trials used IV analgesia as control group. TSA showed that only 22% (133 patients) of the required information size (604 patients) was accrued. The cumulative Z curve did not touch the conventional boundary for benefit or harm (P < 0.05) or the trial sequential monitoring boundary for benefit or harm (Fig. S2). Pneumonia Three trials reported data on pneumonia (n = 108).8,9,11 The conventional meta-analysis showed no statistically significant difference between the intervention group and control group: fixed-effect RR (95% CI) 0.49 (0.19–1.25); P = 0.13; I2 = 0% (Fig. 4). Application of an empirical continuity correction in the no-event trial did not change the result. In the post hoc sensitivity analysis (exclusion of Mackersie et al.11) the point-estimate was identical (zero event trial exclusion). The subgroup analysis of the different comparators showed no statistically significantly increased intervention effect in trials using IV analgesia as Acta Anaesthesiologica Scandinavica (2015)
6
© 2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
EPIDURAL ANALGESIA FOR RIB FRACTURES?
Fig. 3. Analgesic interventions and duration of mechanical ventilation in patients with traumatic rib fractures. Forest plot. CI, confidence interval; IV, intravenous; SD, standard deviation.
control group (test of interaction P = 0.98). TSA could not be conducted due to too few data (DIS < 5%). Secondary outcomes Pain Three trials reported data on pain;6,8,11 however, only two (n = 90) could be included in the metaanalysis and TSA.6,8 The conventional metaanalysis showed no statistically significant difference in pain at rest and in pain at cough/ movement between the CEA group and the control group: random-effects MD (95% CI) −12.71 (−36.71 to 11.36); P = 0.30; I2 = 94% (Fig. S3) and −14.19 (−38.88 to 10.49); P = 0.26; I2 = 97.5% (Fig. S4), respectively. The subgroup analyses of the different comparators showed a statistically significantly increased intervention effect in the paravertebral block trial compared with the intercostal block trial (test of interaction P < 0.0001 and P < 0.00001, respectively). TSA could not be conducted due to too few data (DIS < 5%).
LOS ICU Five trials reported data on ICU LOS (n = 191).6,8,9,24,26 The conventional meta-analysis showed no statistically significant difference in ICU LOS in patients treated with CEA compared with patients treated with another analgesic intervention: random-effects MD (95% CI) −3.49 (−7.54 to 0.55); P = 0.09; I2 = 94% (Fig. S5). In the post hoc sensitivity analysis (exclusion of Ullman et al.24) the point-estimate was: randomeffects MD (95% CI) −0.54 (−1.56 to 0.48); P = 0.26; I2 = 25%. The subgroup analysis of the different comparators showed no statistically significantly increased intervention effect in trials using IV analgesia as control group (test of interaction P = 0.13). TSA showed that only 18% (191 patients) of the required information size (1066 patients) was accrued. The cumulative Z curve did not touch the conventional boundary for benefit or harm (P < 0.05) or the trial sequential monitoring boundary for benefit or harm (Fig. S6).
Acta Anaesthesiologica Scandinavica (2015) © 2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
7
P. DUCH AND M. H. MØLLER
Fig. 4. Analgesic interventions and pneumonia in patients with traumatic rib fractures. Forest plot. CI, confidence interval; M-H, Mantel Haenszel method.
LOS hospital Five trials reported data on hospital LOS (n = 196).6,8,9,11,24 The conventional meta-analysis showed no statistically significant difference in hospital LOS in patients treated with CEA compared with patients treated with another analgesic intervention: random-effects MD (95% CI) −5.72 (−12.23 to 0.78); P = 0.08; I2 = 93% (Fig. S7). In the post hoc sensitivity analysis (exclusion of Mackersie et al. and Ullman et al.11,24) the point-estimate was: fixed-effect MD (95% CI) −1.79 (−3.14 to −0.44); P = 0.66; I2 = 0%. The subgroup analysis of the different comparators showed no statistically significantly increased intervention effect in trials using IV analgesia as control group (test of interaction
P = 0.76). TSA showed that only 19% (196 patients) of the required information size (1053 patients) was accrued. The cumulative Z curve did not touch the conventional boundary for benefit or harm (P < 0.05) or the trial sequential monitoring boundary for benefit or harm (Fig. S8). Pruritus Three trials reported data on pruritus (n = 106).9,11,24 The conventional meta-analysis showed no statistically significant difference in pruritus in patients treated with CEA compared with patients treated with another analgesic intervention: fixed-effect RR (95% CI) 0.86 (0.36– 2.09); P = 0.50; I2 = 0% (Fig. S9). Application of Acta Anaesthesiologica Scandinavica (2015)
8
© 2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
EPIDURAL ANALGESIA FOR RIB FRACTURES?
an empirical continuity correction in the no-event trial did not change the result. In the post hoc sensitivity analysis (exclusion of Mackersie et al. and Ullman et al.11,24) the point-estimate was: fixed-effect RR (95% CI) 1.09 (0.36–3.27); P = 0.88; I2 = 0%. The subgroup analysis of the different comparators could not be conducted because all three trials used IV analgesia as control group. TSA could not be conducted due to too few data (DIS < 5%). Neurological deficits Two trials reported data on neurological deficits (n = 58).8,24 The conventional meta-analysis showed no statistically significant difference in neurological deficits in patients treated with CEA compared with patients treated with another analgesic intervention: fixed-effect RR (95% CI) 0.33 (0.01–7.58); P = 0.49; I2 = 0% (Fig. S10). Application of an empirical continuity correction in the no-event trial did not change the result. In the post hoc sensitivity analysis (exclusion of Ullman et al. 24), the point-estimate was: fixedeffect RR (95% CI) 0.33 (0.01–7.58); P = 0.49; I2 = 0%. The subgroup analysis of the different comparators could not be conducted because of the no-event trial. TSA could not be conducted due to too few data (DIS < 5%). Obstipation, delirium and deep vein thrombosis No trials reported data on these outcome measures. Discussion In the present systematic review of RCTs comparing CEA with other analgesic interventions in patients with traumatic rib fractures, we found no statistically significant difference in terms of clinically relevant patient-centred outcome measures, including mortality, pneumonia, duration of mechanical ventilation and pain. Strengths and limitations of this review The compliance with the recommendations of the Cochrane Collaboration is a strength of the present systematic review, including a preexperimentally published protocol, a systematic literature search with no language restrictions,
independent literature search, data extraction and risk of bias assessment by two authors, contact to authors for further details and the inclusion of trials irrespective of publication status. In addition, we evaluated the risk of random errors with the application of TSA to increase the robustness of the analyses; this methodology has not been used in previous systematic reviews on CEA compared with other analgesic interventions in patients with traumatic rib fractures. We excluded trials not reporting patient-centred outcome measures in order to make the results relevant for clinical practice. We did not define the outcome measures evaluated; rather, we used the definitions proposed by the authors, which may have resulted in trial heterogeneity. Primary outcome measures The conventional meta-analysis showed no benefit of CEA as compared with other analgesic interventions regarding any of the primary outcome measures: mortality, duration of mechanical ventilation and pneumonia. The sensitivity analyses with TSA confirmed this finding. Furthermore, TSA highlighted that less than 22% of the required information size necessary to confirm or reject the null hypothesis has been accrued. All trials had a high risk of bias, indicating poor quality and a risk of falsely inflated point estimates in the cumulative meta-analyses, making the interpretation difficult.15 Considering the high risk of bias and sparse data, no firm evidence for benefit or harm of CEA as compared with other analgesic interventions in terms of mortality, pneumonia and duration mechanical ventilation exists, and considerably, more data are needed. Secondary outcome measures Neither conventional meta-analyses nor TSA showed benefit of CEA as compared with other analgesic interventions regarding any of the secondary outcome measures: ICU and hospital LOS, pain, pruritus and neurological deficits. TSA highlighted the lack of necessary information, as less than 19% of the required information size has been accrued. In the pain subgroup analyses, a statistically significantly increased intervention effect in the paravertebral block trial as compared
Acta Anaesthesiologica Scandinavica (2015) © 2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
9
P. DUCH AND M. H. MØLLER
with the intercostal block trial was noted. However, this finding is based on a very limited number of patients, and as mentioned above, the results of the cumulative meta-analyses may have been influenced by the quality of the included trials.15 The post hoc sensitivity analysis with the exclusion of the two oldest trials using CEA with opioid only11,24 resulted in a statistically significantly shorter duration of mechanical ventilation and hospital LOS in patients treated with CEA compared with patients treated with another analgesic intervention. However, considering the high risk of bias and sparse data, no firm evidence for benefit or harm of CEA as compared with other analgesic interventions in terms of LOS, pain, pruritus and neurological deficits exists, and more data are needed. A limited number of systematic reviews of RCTs comparing CEA with other analgesic interventions in patients with traumatic rib fractures exist. In 2008, Carrier et al. conducted a review comprising eight RCTs.14 In accordance with the findings in the present review, CEA did not significantly affect mortality, ICU LOS, hospital LOS or duration of mechanical ventilation. In 2007 and 2010, respectively, two systematic reviews comprising four RCTs each concluded that based on limited evidence from moderate quality studies, CEA could offer some benefits over IV analgesia.34,35 Finally, a 2003 narrative review of Karmakar and Ho concluded that no specific analgesic modality could be recommended in patients with traumatic rib fractures.36 The present review differs from the reviews above in several ways. First, our review was conducted according to the recommendations by the Cochrane Collaboration, including a systematic literature search, a preexperimentally published protocol and TSA.15 This minimises the risk of systematic errors and random errors, and hereby spurious findings.37 Second, we focused specifically on traumatic rib fractures and not chest trauma in general. Third, we excluded non-randomised trials and quasirandomised trials in order to report non-biased results.38 Fourth, only trials reporting patientcentred outcome measures were included, as surrogate outcome measures overestimate the intervention effect.17 Fifth, we included two new RCTs.6,8 Consequently, we believe that the conclusions of the present systematic review are valid and robust.
Despite the overall clinical impression that CEA is superior to other analgesic interventions in traumatic rib fractures, the present systematic review of RCTs could not confirm it. Existing data have a high risk of bias and are too sparse to conclude on whether CEA is superior to other analgesic interventions in patients with traumatic rib fractures. In everyday clinical practice, it is essential to balance the potential benefit and harm of an intervention. CEA has the potential to be superior to other analgesic interventions in patients with traumatic rib fractures; however, currently, there is no firm evidence for benefit. Furthermore, the potential harm of CEA in patients with traumatic rib fractures, including neurological damage,39,40 urine retention39 and cardiovascular consequences,39 has been scantily investigated. Accordingly, more high-quality research is needed in order to answer the question on the better analgesic intervention in patients with traumatic rib fractures. Conclusion This systematic review of RCTs with metaanalysis and TSA has demonstrated that the quality and quantity of evidence for the use of CEA in patients with traumatic rib fractures is low, and there is no firm evidence for benefit or harm of CEA compared with other analgesic interventions. Well-powered RCTs with low risk of bias reporting clinically relevant patientcentred outcome measures are highly needed. References 1. Gage A, Rivara F, Wang J, Jurkovich GJ, Arbabi S. The effect of epidural placement in patients after blunt thoracic trauma. J Trauma Acute Care Surg 2014; 76: 39–45. 2. Dehghan N, de Mestral C, McKee MD, Schemitsch EH, Nathens A. Flail chest injuries: a review of outcomes and treatment practices from the National Trauma Data Bank. J Trauma Acute Care Surg 2014; 76: 462–8. 3. Flagel BT, Luchette FA, Reed RL, Esposito TJ, Davis KA, Santaniello JM, Gamelli RL. Half-a-dozen ribs: the breakpoint for mortality. Surgery 2005; 138: 717–23. 4. Ziegler DW, Agarwal NN. The morbidity and mortality of rib fractures. J Trauma 1994; 37: 975–9. Acta Anaesthesiologica Scandinavica (2015)
10
© 2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
EPIDURAL ANALGESIA FOR RIB FRACTURES?
5. Bulger EM, Arneson MA, Mock CN, Jurkovich GJ. Rib fractures in the elderly. J Trauma 2000; 48: 1040–6. 6. Hashemzadeh S, Hashemzadeh K, Hosseinzadeh H, Aligholipour Maleki R, Golzari SEJ, Golzari S. Comparison of thoracic epidural and intercostal block to improve ventilation parameters and reduce pain in patients with multiple rib fractures. J Cardiovasc Thorac Res 2011; 3: 87–91. 7. Barnea Y, Kashtan H, Skornick Y, Werbin N. Isolated rib fractures in elderly patients: mortality and morbidity. Can J Surg 2002; 45: 43–6. 8. Mohta M, Verma P, Saxena AK, Sethi AK, Tyagi A, Girotra G. Prospective, randomized comparison of continuous thoracic epidural and thoracic paravertebral infusion in patients with unilateral multiple fractured ribs-a pilot study. J Trauma 2009; 66: 1096–101. 9. Bulger EM, Edwards T, Klotz P, Jurkovich GJ. Epidural analgesia improves outcome after multiple rib fractures. Surgery 2004; 136: 426–30. 10. Moon MR, Luchette FA, Gibson SW, Crews J, Sudarshan G, Hurst JM, Davis K Jr., Johannigman JA, Frame SB, Fischer JE. Prospective, randomized comparison of epidural versus parenteral opioid analgesia in thoracic trauma. Ann Surg 1999; 229: 684–92. 11. Mackersie RC, Karagianes TG, Hoyt DB, Davis JW. Prospective evaluation of epidural and intravenous administration of fentanyl for pain control and restoration of ventilatory function following multiple rib fractures. J Trauma 1991; 31: 443–9. 12. Wu CL, Jani ND, Perkins FM, Barquist E. Thoracic epidural analgesia versus intravenous patient-controlled analgesia for the treatment of rib fracture pain after motor vehicle crash. J Trauma 1999; 47: 564–7. 13. Keus F, Wetterslev J, Gluud C, van Laarhoven CJ. Evidence at a glance: error matrix approach for overviewing available evidence. BMC Med Res Methodol 2010; 10: 90. 14. Carrier FM, Turgeon AF, Nicole PC, Trepanier CA, Fergusson DA, Thauvette D, Lessard MR. Effect of epidural analgesia in patients with traumatic rib fractures: a systematic review and meta-analysis of randomized controlled trials. Can J Anaesth 2009; 56: 230–42. 15. Higgins JPT, Green S (eds). Cochrane handbook for systematic reviews of interventions version 5.1.0 [updated March 2011]. The Cochrane Collaboration 2011. Available at www.cochrane-handbook.org.
16. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. BMJ 2009; 339: b2535. 17. Ciani O, Buyse M, Garside R, Pavay T, Stein K, Sterne JA, Taylor RS. Comparison of treatment effect sizes associated with surrogate and final patient relevant outcomes in randomised controlled trials: meta-epidemiological study. BMJ 2013; 346: f457. 18. Guyatt GH, Oxman AD, Kunz R, Atkins D, Brozek J, Vist G, Alderson P, Glasziou P, Falck-Ytter Y, Schunemann HJ. GRADE guidelines: 2. Framing the question and deciding on important outcomes. J Clin Epidemiol 2011; 64: 395–400. 19. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002; 21: 1539–58. 20. Wetterslev J, Thorlund K, Brok J, Gluud C. Estimating required information size by quantifying diversity in random-effects model meta-analyses. BMC Med Res Methodol 2009; 9: 86. 21. Wetterslev J, Thorlund K, Brok J, Gluud C. Trial sequential analysis may establish when firm evidence is reached in cumulative meta-analysis. J Clin Epidemiol 2008; 61: 64–75. 22. Higgins JP, Whitehead A, Simmonds M. Sequential methods for random-effects meta-analysis. Stat Med 2011; 30: 903–21. 23. Sweeting MJ, Sutton AJ, Lambert PC. What to add to nothing? Use and avoidance of continuity corrections in meta-analysis of sparse data. Stat Med 2004; 23: 1351–75. 24. Ullman DA, Fortune JB, Greenhouse BB, Wimpy RE, Kennedy TM. The treatment of patients with multiple rib fractures using continuous thoracic epidural narcotic infusion. Reg Anesth 1989; 14: 43–7. 25. Copenhagen Trial Unit. Available at: http://www.ctu.dk/tsa/ (accessed 1 October 2014). 26. Sahin S, Uckunkaya N, Soyal S, Haci I. The role of epidural continuous pain treatment on duration of intubation, ventilation and ICU stay in flail chest patients. Agr Derg 1993; 5: 18–20. 27. Hakim SM, Latif FS, Anis SG. Comparison between lumbar and thoracic epidural morphine for severe isolated blunt chest wall trauma: a randomized open-label trial. J Anesth 2012; 26: 836–44. 28. Nejmi H, Fath K, Anaflous R, Sourour S, Samkaoui M. A prospective randomized comparison of nebulized morphine versus thoracic
Acta Anaesthesiologica Scandinavica (2015) © 2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
11
P. DUCH AND M. H. MØLLER
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
epidural analgesia in the management of thoracic trauma. Ann Fr Anesth Reanim 2010; 29: 415–8. Pierre EJ, Martin P, Frohock J, Varon AJ, Barquist E. Lumbar epidural morphine versus patient-controlled analgesia morphine in patients with multiple rib fractures. Anesthesiology 2005; 103: A289. Rousakis G, Fassaris AG, Foniadaki D, Thanopoulos I, Pittakas M, Katsouli-Liapi I. Thoracic epidural versus parenteral analgesia for pain management in multible rib fractures. J Emerg Med 2007; 33: 233–348. Luchette FA, Radafshar SM, Kaiser R, Flynn W, Hassett JM. Prospective evaluation of epidural versus intrapleural catheters for analgesia in chest wall trauma. J Trauma 1994; 36: 865–9. Shinohara K, Iwama H, Akama Y, Tase C. Interpleural block for patients with multiple rib fractures: comparison with epidural block. J Emerg Med 1994; 12: 441–6. Wulf H, Jeckstrom W, Maier CH, Winckler K. Interpleurale Katheteranalgesie bei patienten mit ribbenserienfrakturen. Anaesthesist 1991; 40: 19–24. Jarvis AM, Cook CH, Lindsey DE, Reilley TE, Steinberg SM, Beery PR, Whitmill ML, Papadimos TJ, Stawicki SP. Comparison of epidural versus parenteral analgesia for traumatic rib fractures: a meta-analysis. Scientist 2009; 3: 50–6. Parris R. Towards evidence based emergency medicine: best BETs from the Manchester Royal Infirmary. Epidural analgesia/anaesthesia versus systemic intravenous opioid analgesia in the management of blunt thoracic trauma. Emerg Med J 2007; 24: 848–9. Karmakar MK, Ho AM-H. Acute pain management of patients with multiple fractured ribs. J Trauma 2003; 54: 615–25. Murad MH, Montori VM, Ioannidis JP, Jaeschke R, Devereaux PJ, Prasad K, Neumann I, Carrasco-Labra A, Hatala R, Meade MO, Wyer P, Cook DJ, Guyatt G. How to read a systematic review and meta-analysis and apply the results to patient care. JAMA 2014; 312: 171–9. Savovic´ J, Jones HE, Altman D, Harris R, Juni P, Pildal J, Als-Nielsen B, Balk E, Gluud C, Gluud L, Ioannidis J, Schulz K, Beynon R, Welton N, Wood L, Moher D, Deeks J, Sterne J. Influence of reported study design characteristics on
intervention effect estimates from randomized, controlled trials. Ann Intern Med 2012; 157: 429–38. 39. Baidya DK, Khanna P, Maitra S. Analgesic efficacy and safety of thoracic paravertebral and epidural analgesia for thoracic surgery: a systematic review and meta-analysis. Interact Cardiovasc Thorac Surg 2014; 18: 626–35. 40. Ruppen W, Derry S, McQuay H, Moore RA. Neurologic injury in obstetric patients with epidural analgesia/anesthesia. Anesthesiology 2006; 105: 394–9.
Supporting Information Additional Supporting Information may be found in the online version of this article: Fig. S1. Analgesic interventions and all-cause mortality in patients with traumatic rib fractures. Fig. S2. Trial sequential analysis of duration of mechanical ventilation in patients with traumatic rib fractures. Fig. S3. Analgesic interventions and pain at rest in patients with traumatic rib fractures. Forest plot. Fig. S4. Analgesic interventions and pain at coughing in patients with traumatic rib fractures. Forest plot. Fig. S5. Analgesic interventions and LOS ICU in patients with traumatic rib fractures. Forest plot. Fig. S6. Trial sequential analysis on LOS ICU in patients with traumatic rib fractures. Fig. S7. Analgesic interventions and LOS hospital in patients with traumatic rib fractures. Forest plot. Fig. S8. Trial sequential analysis on LOS hospital in patients with traumatic rib fractures. Fig. S9. Analgesic interventions and pruritus in patients with traumatic rib fractures. Forest plot. Fig. S10. Analgesic interventions and neurological deficit in patients with traumatic rib fractures. Forest plot. Table S1. Characteristics of the excluded trials. Table S2. Summary of findings. Appendix S1. PICO question and search string. Appendix S2. Subgroup analyses. Appendix S3. Trial sequential analysis.
Acta Anaesthesiologica Scandinavica (2015)
12
© 2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
Epidural analgesia in patients with traumatic rib fractures: a systematic review of randomised controlled trials P. Duch1 and M. H. Møller2 1 2
Department of Anaesthesiology and Intensive Care Medicine, Copenhagen University Hospital Hvidovre, Hvidovre, Denmark Department of Intensive Care 4131, Copenhagen University Hospital Rigshospitalet, Copenhagen, Denmark
Correspondence P. Duch, Department of Anaesthesiology and Intensive Care Medicine, Copenhagen University Hospital Hvidovre, Kettegård Alle 39, Hvidovre 2650, Denmark E-mail: [email protected] Conflicts of interest The authors have no conflicts of interest. Funding This research received no grant from any funding agency in the public, commercial or not-for-profit sectors. Submitted 26 December 2014; accepted 29 December 2014; submission 5 November 2014. Citation Duch P, Møller MH. Epidural analgesia compared with other analgesic interventions in patients with traumatic rib fractures. Acta Anaesthesiologica Scandinavica 2015 doi: 10.1111/aas.12475
Background: Traumatic rib fractures are a common condition associated with considerable morbidity and mortality. Observational studies have suggested improved outcome in patients receiving continuous epidural analgesia (CEA). The aim of the present systematic review of randomised controlled trials (RCTs) was to assess the benefit and harm of CEA compared with other analgesic interventions in patients with traumatic rib fractures. Methods: We performed a systematic review with meta-analysis and trial sequential analysis (TSA). Eligible trials were RCTs comparing CEA with other analgesic interventions in patients with traumatic rib fractures. Cumulative relative risks (RRs) and mean differences (MDs) with 95% confidence intervals (CIs) were estimated, and risk of systematic and random errors was assessed. The predefined primary outcome measures were mortality, pneumonia and duration of mechanical ventilation. Results: A total of six trials (n = 223) were included; all were judged as having a high risk of bias. In the conventional metaanalyses, there was no statistically significant difference in mortality (RR 2.18, 95% CI 0.21–22.42; P = 0.51; I2 = 0%), duration of mechanical ventilation (MD −7.53, 95% CI −16.32 to 1.26; P = 0.09; I2 = 91%) or pneumonia (RR 0.49, 95% CI 0.19–1.25; P = 0.13; I2 = 0%) between CEA and other analgesic interventions. Subgroup analyses and sensitivity analyses, including TSA confirmed the results. Conclusion: The quality and quantity of evidence for the use of CEA in patients with traumatic rib fractures is low, and there is no firm evidence for benefit or harm of CEA compared with other analgesic interventions. Well-powered RCTs with low risk of bias reporting clinically relevant patient-centred outcome measures are needed.
Editorial comment: what this article tells us In this well-performed systematic review with meta-analysis and trial sequential analysis, it was somewhat surprisingly demonstrated that there is no firm evidence for benefit or harm of continuous epidural analgesia compared with other analgesic interventions in patients with traumatic rib fractures.
Acta Anaesthesiologica Scandinavica (2015) © 2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
bs_bs_banner
1
P. DUCH AND M. H. MØLLER
Rib fractures are a common condition following trauma with a reported incidence of up to 10% in trauma in general,1 and up to 39% in blunt chest trauma.2 Traumatic rib fractures are associated with significant morbidity and mortality,1 and mortality rates reaching 10–16% has been reported.2,3 An estimated one third of patients with traumatic rib fractures develop secondary pulmonary complications2,4–6 with an associated mortality rate as high as 65%.7 Pain is recognised as a contributing factor to adverse outcome in traumatic rib fractures due to pain-induced inadequate respiratory efforts leading to atelectasis, difficulties in clearing secretions and an increased risk of developing pneumonia.3,6,8,9 Consequently, adequate analgesia is considered a core intervention in the management of patients with traumatic rib fractures.6,10–12 Based on observational data from the United States and Canada and older small randomised controlled trials (RCTs), it has been suggested that continuous epidural analgesia (CEA) is associated with statistically significantly increased survival compared with other analgesic interventions in patients with traumatic rib fractures.1,2,6,9,11 However, it is of major importance to evaluate the risk of design errors, systematic errors (bias) and random errors before concluding on the superiority of one intervention over the other.13 Existing efforts on this matter have been both quantitatively and qualitatively limited.14 Uncertainty over whether CEA is superior to other analgesic interventions in patients with traumatic rib fractures therefore exists among clinicians. Consequently, there is a need to systematically and rigorously re-appraise the better analgesic intervention in patients with traumatic rib fractures. The objective of the present systematic review of RCTs was to evaluate the benefit and harm of CEA compared with other analgesic interventions using the Cochrane Collaboration methodology, including an assessment of the risk of bias and random errors [trial sequential analysis (TSA)].
tematic Reviews and Meta-Analyses statement.16 The protocol has pre-experimentally been published in the International Prospective Register of Systematic Reviews, no. CRD42014010492. Eligibility criteria RCTs in patients with one or more traumatic rib fractures were considered for inclusion, irrespective of language, blinding, publication date and publication status. CEA by placement of an epidural catheter in the epidural space followed by either a repeated bolus, or an infusion of a local anaesthetic agent and/or opioid, was considered the experimental intervention. The control group included one of the following four analgesic interventions: 1. Intravenous (IV) analgesia; bolus or infusion of an opioid carried out by clinical staff (nursecontrolled analgesia) and/or by patient (patient-controlled analgesia). 2. Thoracic paravertebral block; placement of an epidural catheter in the thoracic paravertebral space followed by (repeated) bolus and infusion of a local anaesthetic agent and/or opioid. 3. Interpleural block; placement of an epidural catheter in the pleural cavity followed by (repeated) bolus and infusion of a local anaesthetic agent and/or opioid. 4. Intercostal block; bolus of a local anaesthetic agent and/or opioid in relevant intercostal spaces. No restrictions regarding patient age, or dose or duration of drug administration, in the intervention or control group were applied. Trials were permitted to have more than one control group. Exclusion criteria were trials in animals, trials in patients with non-traumatic rib fractures, quasirandomised trials, observational studies, crossover trials, trials comparing different treatment regimens of CEA and trials not reporting the patient-centred outcome measures17 outlined in the protocol. Outcome measures
Methods This systematic review is based on the methodology recommended by the Cochrane Collaboration,15 and the review has been prepared according to the PRISMA: Preferred Reporting Items for Sys-
The predefined primary outcome measures were: 1. All-cause mortality at longest follow-up. 2. Duration of invasive mechanical ventilation (intermittent positive pressure ventilation). Acta Anaesthesiologica Scandinavica (2015)
2
© 2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
EPIDURAL ANALGESIA FOR RIB FRACTURES?
3. Hospital acquired (nosocomial) pneumonia at longest follow-up (defined according to the definitions used in the included trials). Predefined secondary outcome measures were: 1. Pain (assessed according to the methods used in the included trials, e.g. visual analogue scale). 2. Length of stay in the intensive care unit (LOS ICU). 3. Length of stay in hospital (LOS hospital). 4. Pruritus. 5. Obstipation. 6. Delirium. 7. Neurological deficits. 8. Deep vein thrombosis. Search strategy We framed the following focused research question: ‘Is CEA in patients with traumatic rib fractures superior to other analgesic interventions?’ A population, intervention, comparator and outcomes-based question and literature search was created18 (Supplementary material). The following databases were searched for literature: MEDLINE, including MeSH (January 1966 to July 2014), EMBASE (January 1980 to July 2014) and the Cochrane Library (July 2014). We also hand-searched the reference lists of included trials and other systematic reviews of CEA in traumatic rib fractures. The electronic literature search was last updated 7 July 2014. Study selection We independently reviewed all titles and abstracts identified in the literature search and excluded trials that were obviously not relevant. The remaining trials were evaluated in full text. Disagreements between reviewers were resolved through discussion. Data extraction We independently extracted information from each included trial using a data extraction form. The extracted information included trial characteristics (year of publication, duration, follow-up, country), characteristics of the trial participants (number, inclusion and exclusion criteria), intervention and
control characteristics, and the outcomes of interest. If data in the identified trials were inadequate and/or unclear, we sought to contact the authors for further details. Correspondingly, one author provided additional information.8 Risk of bias assessment In order to evaluate the risk of systematic errors in the included trials, we independently assessed the risk of bias as advised by the Cochrane Collaboration,15 including the domains of random sequence generation, allocation concealment, blinding, incomplete outcome data, selective outcome reporting, baseline imbalance and bias due to vested financial interest. If one or more domains were judged as being high or unclear, we classified the trial as having a high risk of bias.15 Statistical analyses For each included trial, we calculated relative risk (RR) with 95% confidence intervals (CI) for dichotomous outcome measures and the mean difference (MD) with 95% CI for continuous outcome measures, and we pooled these measures in conventional cumulative meta-analyses. Heterogeneity among trials was quantified with inconsistency factor (I2)19 and the diversity (D2) statistics.20 If the I2 statistic was 0, we reported the results from a fixed-effect model, and if the I2 statistic was > 0, we reported results from a random-effects model (the most conservative estimate). The risk of random errors in the cumulative meta-analyses was assessed by TSA20–22 (details on TSA are given in Appendix S3). A sensitivity analysis with application of continuity correction in trials of zero events was conducted.23 In addition, a post hoc sensitivity analysis with exclusion of the two oldest trials using CEA with opioid only were carried out.11,24 Risk of small trial bias by means of funnel plot was abandoned because less than 10 trials were included.15 Review Manager 5.2 (The Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen, Denmark) was used for the conventional meta-analyses, and for the TSA, we used the TSA program version 0.9 beta.25 Subgroup analyses A number of predefined subgroup analyses were planned (Appendix S2); however, only the sub-
Acta Anaesthesiologica Scandinavica (2015) © 2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
3
P. DUCH AND M. H. MØLLER
group analysis of the four comparators could be carried out and is consequently reported.
matter.8,26 None of the trials provided information on duration of follow-up.
Results
Participants
Figure 1 summarises the results of the search. Six trials were included: five published in English6,8,9,11,24 and one published in Turkish26 (Table 1). The main reasons for exclusion of trials were (1) no relevant patient-centred outcome measures were reported;27–30 (2) the population of interest was chest trauma in general and not specifically traumatic rib fractures;10,31 and (3) the design of the trial32,33 (Table S1).
The included trials enrolled a total of 223 patients. Four trials included patients from the trauma centre,8,9,11,26 and two trials included patients from the intensive care unit (ICU).6,24 Inclusion and exclusion criteria were overall comparable between trials (Table 1).
Characteristics of trials All of the included RCTs were single-centred trials. Three were conducted in the United States,9,11,24 one in Iran,6 one in Turkey26 and one in India.8 The duration of inclusion varied from 17 to 46 months; two trials did not report on this n = 336 of records identified through database (MEDLINE) searching after duplicates removed
Intervention and comparators Three of the six trials evaluated CEA with a local anaesthetic agent and an opioid,6,9,26 whereas one trial evaluated CEA without additional opioid,8 and two trials evaluated CEA with opioid alone.11,24 The comparator was IV analgesia in four trials,9,11,24,26 paravertebral block in one trial8 and intercostal block in one trial.6 None of the included trials used interpleural block as comparator. n = 4 of additional records identified through hand-search of the reference lists and other sources according to our search strategy
n = 340 of records screened on basis of title and abstract
n = 326 of records excluded
n = 14 of full-text articles assessed for eligibility
n = 8 of full-text articles excluded: Not RCT (n = 1)
n = 6 of studies included in qualitative synthesis
n = 6 of studies included in quantitative synthesis (metaanalysis and trial sequential analysis)
Cross-over design (n = 1) Not relevant end point (n = 1) Chest trauma in general (not rib fractures) (n = 2) Inadequately reported end points (n = 3)
Fig. 1. Study flow diagram. RCT, randomised controlled trial. Acta Anaesthesiologica Scandinavica (2015)
4
© 2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
ICU
60
Hascehmzadeh et al. 20116
© 2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
Trauma/single India centre
Trauma/single Turkey centre
Surgical ICU
30
27
28
Mohta et al. 20098
Sahin et al. 199326
Ullman et al. 198924
36
Unknown
Unknown
36
17
46
3 or more unilateral rib Not mentioned fractures or flail segment with significant contusion of the chest wall with impaired ventilation. Injuries had to be associated with significant pain and splinting with normal ventilation.
Traumatic brain injury.
Severe altered mental status. Unstable cardiac status. Liver or kidney disease. Spine injury. Pelvic fracture. Abdominal visceral injuries. Traumatic brain injury.
> 18 years > 3 unilateral fractured ribs.
Flail chest
Pregnancy. History of substance abuse. Chronic use of analgesics. Chronic pain. Spine injury. Painful limp injury.
> 18 years > 3 rib fractures and flail chest or > 2 rib fractures and exploratory laparotomy/ pulmonary contusion. Normal mental status
Meperidin IV, PCA
Paravertebral catheter with continuous bupivacain and IV morphine p.r.n.
Continuous IV fentanyl infusion
Intercostal block every 8 h with bupivacain. IV pethidine p.r.n.
IV opioids: Morphine, hydromorphine and fentanyl. Administered by patient-controlled analgesia (PCA) for alert patients and with nurse assistance (NCA) for patients who could not participate in self-administration
Comparator
Thoracic epidural catheter Continuous IV morphine or IV with loading dose morphine morphine every 3–4 hours and fentanyl and p.r.n. continuous morphine.
Epidural catheter with continuous bupivacain and fentanyl infusion
Thoracal epidural catheter with continuous bupivacaine and IV morphine p.r.n.
Lumbar epidural catheter with continuous fentanyl infusion
Thoracic epidural catheter with bupivacain and morphine every 8 h. And IV pethidine p.r.n.
Cerebral injury. Mechanical ventilation. Systemic infection. Fever. Liver or splenic trauma.
> 18 years > 1 rib fracture GCS > 14
Intervention
Spine injury. Severe traumatic brain Epidural catheter with injury. Severe altered mental continuous bupivacaine, status. Unstable pelvic fracture or morphine, and fentanyl open abdomen that would preclude positioning for epidural placement. Cardiac instability or coagulopathy. Active chest wall infection. Acute thoracic aortic transection. Patients whose chest wall pain was manageable with oral opioids or anti-inflammatory medications.
Exclusion criteria (Population)
> 18 years 3 or more rib fractures
GCS, Glasgow Coma Score; ICU, intensive care unit; LOS, length of stay; p.r.n., pro re nata.
USA
Trauma/single USA centre
Mackersie et al. 32 199111
Iran
Trauma/single USA centre
46
Duration of inclusion Inclusion criteria Country (months) (Population)
Bulger et al. 20049
Number of patients in trial Setting
Table 1 Characteristics of the included trials.
Mechanical ventilation LOS ICU LOS Hospital Pruritus Neurological deficit Duration of follow-up unknown
Mechanical ventilation LOS ICU Duration of follow-up unknown
Mortality Pneumonia LOS ICU LOS Hospital Pain Neurological deficit Duration of follow-up unknown
Pneumonia Mechanical ventilation LOS Hospital Pain Pruritus Duration of follow-up unknown
LOS ICU LOS Hospital Pain Duration of follow-up unknown
Mortality Pneumonia Mechanical ventilation LOS ICU LOS Hospital Pruritus Duration of follow-up unknown
Outcomes
EPIDURAL ANALGESIA FOR RIB FRACTURES?
Acta Anaesthesiologica Scandinavica (2015)
5
P. DUCH AND M. H. MØLLER
analysis showed no significant difference in mortality in patients treated with CEA compared with those treated with another analgesic intervention: fixed-effect RR (95% CI) 2.18 (0.21–22.42); P = 0.51; I2 = 0% (Fig. S1). Application of an empirical continuity correction in the no-event trial did not change the result. The subgroup analysis of trials using IV analgesia as comparator versus trials using paravertebral block as comparator could not be conducted because of the no-event trial. TSA could not be conducted due to too few data [diversity (D2) adjusted required information size (DIS) < 5%].
Fig. 2. Risk of bias summary.
Risk of bias No trials were judged to be of low risk of bias in all six domains (Fig. 2). The main reasons for high risk of bias were lack of blinding of patients and personnel, and lack of blinding of outcome assessment. Four trials had high risk of reporting bias, because they did not include results for the key outcome measure mortality.6,11,24,26 One trial had inadequate random sequence generation and allocation concealment,24 and one trial had high risk of other bias because of crossover between intervention and control group.9
Outcome measures All-cause mortality Mortality data were obtained from two trials including 76 patients.8,9 The conventional meta-
Duration of mechanical ventilation Four trials (n = 133) reported data on duration of mechanical ventilation.9,11,24,26 The conventional meta-analysis showed no statistically significant difference in duration of mechanical ventilation in patients treated with CEA compared with patients treated with another analgesic intervention: random-effects MD (95% CI) −7.53 (−16.32 to 1.26); P = 0.09; I2 = 91% (Fig. 3). In the post hoc sensitivity analysis (exclusion of Mackersie et al. and Ullman et al.11,24), the point-estimate was: fixed-effect MD (95% CI) −4.17 (−5.45 to −2.88); P = 0.61; I2 = 0%. The subgroup analysis of the different comparators could not be conducted because all four trials used IV analgesia as control group. TSA showed that only 22% (133 patients) of the required information size (604 patients) was accrued. The cumulative Z curve did not touch the conventional boundary for benefit or harm (P < 0.05) or the trial sequential monitoring boundary for benefit or harm (Fig. S2). Pneumonia Three trials reported data on pneumonia (n = 108).8,9,11 The conventional meta-analysis showed no statistically significant difference between the intervention group and control group: fixed-effect RR (95% CI) 0.49 (0.19–1.25); P = 0.13; I2 = 0% (Fig. 4). Application of an empirical continuity correction in the no-event trial did not change the result. In the post hoc sensitivity analysis (exclusion of Mackersie et al.11) the point-estimate was identical (zero event trial exclusion). The subgroup analysis of the different comparators showed no statistically significantly increased intervention effect in trials using IV analgesia as Acta Anaesthesiologica Scandinavica (2015)
6
© 2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
EPIDURAL ANALGESIA FOR RIB FRACTURES?
Fig. 3. Analgesic interventions and duration of mechanical ventilation in patients with traumatic rib fractures. Forest plot. CI, confidence interval; IV, intravenous; SD, standard deviation.
control group (test of interaction P = 0.98). TSA could not be conducted due to too few data (DIS < 5%). Secondary outcomes Pain Three trials reported data on pain;6,8,11 however, only two (n = 90) could be included in the metaanalysis and TSA.6,8 The conventional metaanalysis showed no statistically significant difference in pain at rest and in pain at cough/ movement between the CEA group and the control group: random-effects MD (95% CI) −12.71 (−36.71 to 11.36); P = 0.30; I2 = 94% (Fig. S3) and −14.19 (−38.88 to 10.49); P = 0.26; I2 = 97.5% (Fig. S4), respectively. The subgroup analyses of the different comparators showed a statistically significantly increased intervention effect in the paravertebral block trial compared with the intercostal block trial (test of interaction P < 0.0001 and P < 0.00001, respectively). TSA could not be conducted due to too few data (DIS < 5%).
LOS ICU Five trials reported data on ICU LOS (n = 191).6,8,9,24,26 The conventional meta-analysis showed no statistically significant difference in ICU LOS in patients treated with CEA compared with patients treated with another analgesic intervention: random-effects MD (95% CI) −3.49 (−7.54 to 0.55); P = 0.09; I2 = 94% (Fig. S5). In the post hoc sensitivity analysis (exclusion of Ullman et al.24) the point-estimate was: randomeffects MD (95% CI) −0.54 (−1.56 to 0.48); P = 0.26; I2 = 25%. The subgroup analysis of the different comparators showed no statistically significantly increased intervention effect in trials using IV analgesia as control group (test of interaction P = 0.13). TSA showed that only 18% (191 patients) of the required information size (1066 patients) was accrued. The cumulative Z curve did not touch the conventional boundary for benefit or harm (P < 0.05) or the trial sequential monitoring boundary for benefit or harm (Fig. S6).
Acta Anaesthesiologica Scandinavica (2015) © 2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
7
P. DUCH AND M. H. MØLLER
Fig. 4. Analgesic interventions and pneumonia in patients with traumatic rib fractures. Forest plot. CI, confidence interval; M-H, Mantel Haenszel method.
LOS hospital Five trials reported data on hospital LOS (n = 196).6,8,9,11,24 The conventional meta-analysis showed no statistically significant difference in hospital LOS in patients treated with CEA compared with patients treated with another analgesic intervention: random-effects MD (95% CI) −5.72 (−12.23 to 0.78); P = 0.08; I2 = 93% (Fig. S7). In the post hoc sensitivity analysis (exclusion of Mackersie et al. and Ullman et al.11,24) the point-estimate was: fixed-effect MD (95% CI) −1.79 (−3.14 to −0.44); P = 0.66; I2 = 0%. The subgroup analysis of the different comparators showed no statistically significantly increased intervention effect in trials using IV analgesia as control group (test of interaction
P = 0.76). TSA showed that only 19% (196 patients) of the required information size (1053 patients) was accrued. The cumulative Z curve did not touch the conventional boundary for benefit or harm (P < 0.05) or the trial sequential monitoring boundary for benefit or harm (Fig. S8). Pruritus Three trials reported data on pruritus (n = 106).9,11,24 The conventional meta-analysis showed no statistically significant difference in pruritus in patients treated with CEA compared with patients treated with another analgesic intervention: fixed-effect RR (95% CI) 0.86 (0.36– 2.09); P = 0.50; I2 = 0% (Fig. S9). Application of Acta Anaesthesiologica Scandinavica (2015)
8
© 2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
EPIDURAL ANALGESIA FOR RIB FRACTURES?
an empirical continuity correction in the no-event trial did not change the result. In the post hoc sensitivity analysis (exclusion of Mackersie et al. and Ullman et al.11,24) the point-estimate was: fixed-effect RR (95% CI) 1.09 (0.36–3.27); P = 0.88; I2 = 0%. The subgroup analysis of the different comparators could not be conducted because all three trials used IV analgesia as control group. TSA could not be conducted due to too few data (DIS < 5%). Neurological deficits Two trials reported data on neurological deficits (n = 58).8,24 The conventional meta-analysis showed no statistically significant difference in neurological deficits in patients treated with CEA compared with patients treated with another analgesic intervention: fixed-effect RR (95% CI) 0.33 (0.01–7.58); P = 0.49; I2 = 0% (Fig. S10). Application of an empirical continuity correction in the no-event trial did not change the result. In the post hoc sensitivity analysis (exclusion of Ullman et al. 24), the point-estimate was: fixedeffect RR (95% CI) 0.33 (0.01–7.58); P = 0.49; I2 = 0%. The subgroup analysis of the different comparators could not be conducted because of the no-event trial. TSA could not be conducted due to too few data (DIS < 5%). Obstipation, delirium and deep vein thrombosis No trials reported data on these outcome measures. Discussion In the present systematic review of RCTs comparing CEA with other analgesic interventions in patients with traumatic rib fractures, we found no statistically significant difference in terms of clinically relevant patient-centred outcome measures, including mortality, pneumonia, duration of mechanical ventilation and pain. Strengths and limitations of this review The compliance with the recommendations of the Cochrane Collaboration is a strength of the present systematic review, including a preexperimentally published protocol, a systematic literature search with no language restrictions,
independent literature search, data extraction and risk of bias assessment by two authors, contact to authors for further details and the inclusion of trials irrespective of publication status. In addition, we evaluated the risk of random errors with the application of TSA to increase the robustness of the analyses; this methodology has not been used in previous systematic reviews on CEA compared with other analgesic interventions in patients with traumatic rib fractures. We excluded trials not reporting patient-centred outcome measures in order to make the results relevant for clinical practice. We did not define the outcome measures evaluated; rather, we used the definitions proposed by the authors, which may have resulted in trial heterogeneity. Primary outcome measures The conventional meta-analysis showed no benefit of CEA as compared with other analgesic interventions regarding any of the primary outcome measures: mortality, duration of mechanical ventilation and pneumonia. The sensitivity analyses with TSA confirmed this finding. Furthermore, TSA highlighted that less than 22% of the required information size necessary to confirm or reject the null hypothesis has been accrued. All trials had a high risk of bias, indicating poor quality and a risk of falsely inflated point estimates in the cumulative meta-analyses, making the interpretation difficult.15 Considering the high risk of bias and sparse data, no firm evidence for benefit or harm of CEA as compared with other analgesic interventions in terms of mortality, pneumonia and duration mechanical ventilation exists, and considerably, more data are needed. Secondary outcome measures Neither conventional meta-analyses nor TSA showed benefit of CEA as compared with other analgesic interventions regarding any of the secondary outcome measures: ICU and hospital LOS, pain, pruritus and neurological deficits. TSA highlighted the lack of necessary information, as less than 19% of the required information size has been accrued. In the pain subgroup analyses, a statistically significantly increased intervention effect in the paravertebral block trial as compared
Acta Anaesthesiologica Scandinavica (2015) © 2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
9
P. DUCH AND M. H. MØLLER
with the intercostal block trial was noted. However, this finding is based on a very limited number of patients, and as mentioned above, the results of the cumulative meta-analyses may have been influenced by the quality of the included trials.15 The post hoc sensitivity analysis with the exclusion of the two oldest trials using CEA with opioid only11,24 resulted in a statistically significantly shorter duration of mechanical ventilation and hospital LOS in patients treated with CEA compared with patients treated with another analgesic intervention. However, considering the high risk of bias and sparse data, no firm evidence for benefit or harm of CEA as compared with other analgesic interventions in terms of LOS, pain, pruritus and neurological deficits exists, and more data are needed. A limited number of systematic reviews of RCTs comparing CEA with other analgesic interventions in patients with traumatic rib fractures exist. In 2008, Carrier et al. conducted a review comprising eight RCTs.14 In accordance with the findings in the present review, CEA did not significantly affect mortality, ICU LOS, hospital LOS or duration of mechanical ventilation. In 2007 and 2010, respectively, two systematic reviews comprising four RCTs each concluded that based on limited evidence from moderate quality studies, CEA could offer some benefits over IV analgesia.34,35 Finally, a 2003 narrative review of Karmakar and Ho concluded that no specific analgesic modality could be recommended in patients with traumatic rib fractures.36 The present review differs from the reviews above in several ways. First, our review was conducted according to the recommendations by the Cochrane Collaboration, including a systematic literature search, a preexperimentally published protocol and TSA.15 This minimises the risk of systematic errors and random errors, and hereby spurious findings.37 Second, we focused specifically on traumatic rib fractures and not chest trauma in general. Third, we excluded non-randomised trials and quasirandomised trials in order to report non-biased results.38 Fourth, only trials reporting patientcentred outcome measures were included, as surrogate outcome measures overestimate the intervention effect.17 Fifth, we included two new RCTs.6,8 Consequently, we believe that the conclusions of the present systematic review are valid and robust.
Despite the overall clinical impression that CEA is superior to other analgesic interventions in traumatic rib fractures, the present systematic review of RCTs could not confirm it. Existing data have a high risk of bias and are too sparse to conclude on whether CEA is superior to other analgesic interventions in patients with traumatic rib fractures. In everyday clinical practice, it is essential to balance the potential benefit and harm of an intervention. CEA has the potential to be superior to other analgesic interventions in patients with traumatic rib fractures; however, currently, there is no firm evidence for benefit. Furthermore, the potential harm of CEA in patients with traumatic rib fractures, including neurological damage,39,40 urine retention39 and cardiovascular consequences,39 has been scantily investigated. Accordingly, more high-quality research is needed in order to answer the question on the better analgesic intervention in patients with traumatic rib fractures. Conclusion This systematic review of RCTs with metaanalysis and TSA has demonstrated that the quality and quantity of evidence for the use of CEA in patients with traumatic rib fractures is low, and there is no firm evidence for benefit or harm of CEA compared with other analgesic interventions. Well-powered RCTs with low risk of bias reporting clinically relevant patientcentred outcome measures are highly needed. References 1. Gage A, Rivara F, Wang J, Jurkovich GJ, Arbabi S. The effect of epidural placement in patients after blunt thoracic trauma. J Trauma Acute Care Surg 2014; 76: 39–45. 2. Dehghan N, de Mestral C, McKee MD, Schemitsch EH, Nathens A. Flail chest injuries: a review of outcomes and treatment practices from the National Trauma Data Bank. J Trauma Acute Care Surg 2014; 76: 462–8. 3. Flagel BT, Luchette FA, Reed RL, Esposito TJ, Davis KA, Santaniello JM, Gamelli RL. Half-a-dozen ribs: the breakpoint for mortality. Surgery 2005; 138: 717–23. 4. Ziegler DW, Agarwal NN. The morbidity and mortality of rib fractures. J Trauma 1994; 37: 975–9. Acta Anaesthesiologica Scandinavica (2015)
10
© 2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
EPIDURAL ANALGESIA FOR RIB FRACTURES?
5. Bulger EM, Arneson MA, Mock CN, Jurkovich GJ. Rib fractures in the elderly. J Trauma 2000; 48: 1040–6. 6. Hashemzadeh S, Hashemzadeh K, Hosseinzadeh H, Aligholipour Maleki R, Golzari SEJ, Golzari S. Comparison of thoracic epidural and intercostal block to improve ventilation parameters and reduce pain in patients with multiple rib fractures. J Cardiovasc Thorac Res 2011; 3: 87–91. 7. Barnea Y, Kashtan H, Skornick Y, Werbin N. Isolated rib fractures in elderly patients: mortality and morbidity. Can J Surg 2002; 45: 43–6. 8. Mohta M, Verma P, Saxena AK, Sethi AK, Tyagi A, Girotra G. Prospective, randomized comparison of continuous thoracic epidural and thoracic paravertebral infusion in patients with unilateral multiple fractured ribs-a pilot study. J Trauma 2009; 66: 1096–101. 9. Bulger EM, Edwards T, Klotz P, Jurkovich GJ. Epidural analgesia improves outcome after multiple rib fractures. Surgery 2004; 136: 426–30. 10. Moon MR, Luchette FA, Gibson SW, Crews J, Sudarshan G, Hurst JM, Davis K Jr., Johannigman JA, Frame SB, Fischer JE. Prospective, randomized comparison of epidural versus parenteral opioid analgesia in thoracic trauma. Ann Surg 1999; 229: 684–92. 11. Mackersie RC, Karagianes TG, Hoyt DB, Davis JW. Prospective evaluation of epidural and intravenous administration of fentanyl for pain control and restoration of ventilatory function following multiple rib fractures. J Trauma 1991; 31: 443–9. 12. Wu CL, Jani ND, Perkins FM, Barquist E. Thoracic epidural analgesia versus intravenous patient-controlled analgesia for the treatment of rib fracture pain after motor vehicle crash. J Trauma 1999; 47: 564–7. 13. Keus F, Wetterslev J, Gluud C, van Laarhoven CJ. Evidence at a glance: error matrix approach for overviewing available evidence. BMC Med Res Methodol 2010; 10: 90. 14. Carrier FM, Turgeon AF, Nicole PC, Trepanier CA, Fergusson DA, Thauvette D, Lessard MR. Effect of epidural analgesia in patients with traumatic rib fractures: a systematic review and meta-analysis of randomized controlled trials. Can J Anaesth 2009; 56: 230–42. 15. Higgins JPT, Green S (eds). Cochrane handbook for systematic reviews of interventions version 5.1.0 [updated March 2011]. The Cochrane Collaboration 2011. Available at www.cochrane-handbook.org.
16. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. BMJ 2009; 339: b2535. 17. Ciani O, Buyse M, Garside R, Pavay T, Stein K, Sterne JA, Taylor RS. Comparison of treatment effect sizes associated with surrogate and final patient relevant outcomes in randomised controlled trials: meta-epidemiological study. BMJ 2013; 346: f457. 18. Guyatt GH, Oxman AD, Kunz R, Atkins D, Brozek J, Vist G, Alderson P, Glasziou P, Falck-Ytter Y, Schunemann HJ. GRADE guidelines: 2. Framing the question and deciding on important outcomes. J Clin Epidemiol 2011; 64: 395–400. 19. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002; 21: 1539–58. 20. Wetterslev J, Thorlund K, Brok J, Gluud C. Estimating required information size by quantifying diversity in random-effects model meta-analyses. BMC Med Res Methodol 2009; 9: 86. 21. Wetterslev J, Thorlund K, Brok J, Gluud C. Trial sequential analysis may establish when firm evidence is reached in cumulative meta-analysis. J Clin Epidemiol 2008; 61: 64–75. 22. Higgins JP, Whitehead A, Simmonds M. Sequential methods for random-effects meta-analysis. Stat Med 2011; 30: 903–21. 23. Sweeting MJ, Sutton AJ, Lambert PC. What to add to nothing? Use and avoidance of continuity corrections in meta-analysis of sparse data. Stat Med 2004; 23: 1351–75. 24. Ullman DA, Fortune JB, Greenhouse BB, Wimpy RE, Kennedy TM. The treatment of patients with multiple rib fractures using continuous thoracic epidural narcotic infusion. Reg Anesth 1989; 14: 43–7. 25. Copenhagen Trial Unit. Available at: http://www.ctu.dk/tsa/ (accessed 1 October 2014). 26. Sahin S, Uckunkaya N, Soyal S, Haci I. The role of epidural continuous pain treatment on duration of intubation, ventilation and ICU stay in flail chest patients. Agr Derg 1993; 5: 18–20. 27. Hakim SM, Latif FS, Anis SG. Comparison between lumbar and thoracic epidural morphine for severe isolated blunt chest wall trauma: a randomized open-label trial. J Anesth 2012; 26: 836–44. 28. Nejmi H, Fath K, Anaflous R, Sourour S, Samkaoui M. A prospective randomized comparison of nebulized morphine versus thoracic
Acta Anaesthesiologica Scandinavica (2015) © 2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
11
P. DUCH AND M. H. MØLLER
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
epidural analgesia in the management of thoracic trauma. Ann Fr Anesth Reanim 2010; 29: 415–8. Pierre EJ, Martin P, Frohock J, Varon AJ, Barquist E. Lumbar epidural morphine versus patient-controlled analgesia morphine in patients with multiple rib fractures. Anesthesiology 2005; 103: A289. Rousakis G, Fassaris AG, Foniadaki D, Thanopoulos I, Pittakas M, Katsouli-Liapi I. Thoracic epidural versus parenteral analgesia for pain management in multible rib fractures. J Emerg Med 2007; 33: 233–348. Luchette FA, Radafshar SM, Kaiser R, Flynn W, Hassett JM. Prospective evaluation of epidural versus intrapleural catheters for analgesia in chest wall trauma. J Trauma 1994; 36: 865–9. Shinohara K, Iwama H, Akama Y, Tase C. Interpleural block for patients with multiple rib fractures: comparison with epidural block. J Emerg Med 1994; 12: 441–6. Wulf H, Jeckstrom W, Maier CH, Winckler K. Interpleurale Katheteranalgesie bei patienten mit ribbenserienfrakturen. Anaesthesist 1991; 40: 19–24. Jarvis AM, Cook CH, Lindsey DE, Reilley TE, Steinberg SM, Beery PR, Whitmill ML, Papadimos TJ, Stawicki SP. Comparison of epidural versus parenteral analgesia for traumatic rib fractures: a meta-analysis. Scientist 2009; 3: 50–6. Parris R. Towards evidence based emergency medicine: best BETs from the Manchester Royal Infirmary. Epidural analgesia/anaesthesia versus systemic intravenous opioid analgesia in the management of blunt thoracic trauma. Emerg Med J 2007; 24: 848–9. Karmakar MK, Ho AM-H. Acute pain management of patients with multiple fractured ribs. J Trauma 2003; 54: 615–25. Murad MH, Montori VM, Ioannidis JP, Jaeschke R, Devereaux PJ, Prasad K, Neumann I, Carrasco-Labra A, Hatala R, Meade MO, Wyer P, Cook DJ, Guyatt G. How to read a systematic review and meta-analysis and apply the results to patient care. JAMA 2014; 312: 171–9. Savovic´ J, Jones HE, Altman D, Harris R, Juni P, Pildal J, Als-Nielsen B, Balk E, Gluud C, Gluud L, Ioannidis J, Schulz K, Beynon R, Welton N, Wood L, Moher D, Deeks J, Sterne J. Influence of reported study design characteristics on
intervention effect estimates from randomized, controlled trials. Ann Intern Med 2012; 157: 429–38. 39. Baidya DK, Khanna P, Maitra S. Analgesic efficacy and safety of thoracic paravertebral and epidural analgesia for thoracic surgery: a systematic review and meta-analysis. Interact Cardiovasc Thorac Surg 2014; 18: 626–35. 40. Ruppen W, Derry S, McQuay H, Moore RA. Neurologic injury in obstetric patients with epidural analgesia/anesthesia. Anesthesiology 2006; 105: 394–9.
Supporting Information Additional Supporting Information may be found in the online version of this article: Fig. S1. Analgesic interventions and all-cause mortality in patients with traumatic rib fractures. Fig. S2. Trial sequential analysis of duration of mechanical ventilation in patients with traumatic rib fractures. Fig. S3. Analgesic interventions and pain at rest in patients with traumatic rib fractures. Forest plot. Fig. S4. Analgesic interventions and pain at coughing in patients with traumatic rib fractures. Forest plot. Fig. S5. Analgesic interventions and LOS ICU in patients with traumatic rib fractures. Forest plot. Fig. S6. Trial sequential analysis on LOS ICU in patients with traumatic rib fractures. Fig. S7. Analgesic interventions and LOS hospital in patients with traumatic rib fractures. Forest plot. Fig. S8. Trial sequential analysis on LOS hospital in patients with traumatic rib fractures. Fig. S9. Analgesic interventions and pruritus in patients with traumatic rib fractures. Forest plot. Fig. S10. Analgesic interventions and neurological deficit in patients with traumatic rib fractures. Forest plot. Table S1. Characteristics of the excluded trials. Table S2. Summary of findings. Appendix S1. PICO question and search string. Appendix S2. Subgroup analyses. Appendix S3. Trial sequential analysis.
Acta Anaesthesiologica Scandinavica (2015)
12
© 2015 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
