Tramadol as adjunct to psoas compartment block with levobupivacaine 0.5%: a randomized double-blinded study{dagger}

S. Mannion*, S. O'Callaghan, D. B. Murphy and G. D. Shorten

Department of Anaesthesia and Intensive Care, Cork University Hospital, St Mary's Orthopaedic Hospital and University College Cork, Cork, Ireland{ddagger}

* Corresponding author: Department of Anaesthesia and Intensive Care, Cork University Hospital, Cork, Ireland. E-mail: mannionstephen{at}hotmail.com

Accepted for publication November 8, 2004.


    Abstract
 Top
 Footnotes
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Background. Tramadol has been administered peripherally to prolong analgesia after brachial plexus and neuraxial blocks. Our aim was to evaluate the systemic and perineural effects of tramadol as an analgesic adjunct to psoas compartment block (PCB) with levobupivacaine.

Methods. In a randomized, prospective, double-blinded trial, 60 patients (ASA I–III), aged 49–88 yr, undergoing primary total hip or knee arthroplasty underwent PCB and subsequent bupivacaine spinal anaesthesia. Patients were randomized into three groups. Each patient received PCB with levobupivacaine 0.5%, 0.4 ml kg–1. The control group (group L, n=21) received i.v. saline, the systemic tramadol group (group IT, n=19) received i.v. tramadol 1.5 mg kg–1 and the perineural tramadol group (group T, n=20) received i.v. saline and PCB with tramadol 1.5 mg kg–1. Postoperatively patients received regular paracetamol 6-hourly and diclofenac sodium 12-hourly. Time to first morphine analgesia, 24-hour morphine consumption, sensory block, pain and sedation scores and haemodynamic parameters were recorded.

Results. Time (h) to first morphine analgesia was similar in the three groups [mean (SD)]: group L, 11.2 (6.6); group T, 14.5 (8.0); group IT, 14.6 (6.8); P=0.35. Twenty-four-hour cumulative morphine (mg) consumption was also similar in the three groups [group L, 21.9 (10.1); group T, 19.8 (6.7), group IT, 16.5 (9.5)], as were durations of sensory and motor block. There were no differences in the incidence of adverse effects except that patients in group IT were more sedated at 14 h than group L (P=0.02).

Conclusion. We conclude that our data do not support a clinically important local anaesthetic or peripheral analgesic effect of tramadol as adjunct to PCB with levobupivacaine 0.5%.

Keywords: anaesthetic techniques, regional, psoas compartment block ; analgesics opioid, tramadol ; pain, postoperative ; surgery, orthopaedic


    Introduction
 Top
 Footnotes
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Clonidine and other adjunctive agents are commonly administered to prolong regional blocks.1 Although clonidine successfully prolongs brachial plexus block,2 concerns regarding its adverse haemodynamic3 and sedative effects4 have prompted investigators to examine alternative adjuncts for regional anaesthesia.

Tramadol is an analgesic drug, acting at central and peripheral µ-opioid and monoaminergic receptors.5 Tramadol also exhibits local anaesthetic properties.6 A study investigating the effects of tramadol 100 mg, sufentanil 20 µg and clonidine 1.5 µg kg–1, combined with ropivacaine 0.75%, demonstrated that each adjunct similarly prolonged analgesia and anaesthesia of brachial plexus block, but that the incidence of adverse effects was less in the tramadol than clonidine group.4

The use of tramadol as an adjunct for brachial plexus block is supported by two further studies.7 8 However, its use in other regional block settings has produced less consistent results.9 10

The use of adjunctive tramadol has not been investigated in the setting of the psoas compartment block (PCB). Despite clonidine's beneficial effects in brachial plexus blockade, we have recently shown that perineural clonidine (1 µg kg–1) does not prolong the action of levobupivacaine 0.5% in PCB.11

The aims of our study were therefore (i) to determine if tramadol (1.5 mg kg–1) prolongs analgesia after PCB with levobupivacaine 0.5% and (ii) to compare the analgesic effect of tramadol after i.v. administration compared with perineural administration with levobupivacaine.


    Methods
 Top
 Footnotes
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
After institutional ethical committee approval and having obtained written informed consent from each patient, 60 ASA physical status I–III patients, aged 49–88 yr, scheduled for either primary hip or knee arthroplasty were included in a prospective, double-blinded, randomized, controlled trial. Exclusion criteria were concurrent or chronic medication with tramadol, perioperative use of ondansetron and contraindications to regional anaesthesia or NSAIDs.

On arrival of the patient in the anaesthesia induction room, a 16–18 G i.v. cannula was inserted. Routine monitoring consisted of continuous electrocardiography, pulse oximetry and non-invasive blood pressure at 5-min intervals (AS 3; Datex Instrumenterium, Finland). Baseline blood pressure was recorded. Patients then received midazolam 2–5 mg i.v. before PCB and Hartmann's solution 10 ml kg–1 was rapidly administered. All patients underwent PCB with subsequent spinal anaesthesia.

Patients were randomly assigned to one of three groups according to randomization tables. The results were made available to investigators using sealed envelopes, one for each patient recruited. Patients in group L received levobupivacaine 0.5%, 0.4 ml kg–1 (Chirocaine®; Abbott Laboratories, Dublin, Ireland) for PCB and saline 0.9%, 5 ml i.v. Patients in group T received levobupivacaine 0.4 ml kg–1, 0.5% in combination with tramadol 1.5 mg kg–1 (50 mg ml–1) (Tradol®; Rowex, Bantry, Ireland) for PCB and saline 0.9%, 5 ml i.v. Patients in group IT received levobupivacaine 0.5%, 0.4 ml kg–1 for PCB and tramadol 1.5 mg kg–1 i.v. made up to 5 ml with saline 0.9%. The drug solutions to be administered were prepared by an anaesthetist not involved in block performance, patient care or data collection.

The same operator (SM) performed all PCBs. The patients were placed in the lateral position, operative side uppermost. The technique for PCB was as described by Capdevila and colleagues.12 After identification of the skin point for needle insertion, lidocaine 1%, 3 ml was injected into the skin and subcutaneous tissues. The skin was prepared with chlorhexidine. A Stimuplex A® 100 mm needle (B. Braun Medical, Melsungen, Germany) was inserted using a nerve stimulator (Stimuplex S®; B. Braun Medical) with a starting output of 1.5 mA and 2 Hz. The needle was advanced until quadriceps twitches were elicited with a current between 0.3 and 0.5 mA. After negative aspiration for blood, the local anaesthetic solution was injected over ~3 min in 5 ml increments. The time at which all of the local anaesthetic solution had been injected was taken as time zero. Patients remained in this position and the i.v. solution was administered.

Sensory blockade of the femoral nerve was assessed 5 and 7.5 min after time zero. Sensory testing for thermoanaesthesia was performed on the mid-anterior thigh with ethyl chloride spray and compared with the contralateral side. Sensation was scored as 0=no difference, 1=less cold, 2=not cold. Sensory onset was defined as complete thermoanaesthesia (i.e. score=2). If no evidence of femoral sensory block was detectable at 7.5 min, the anaesthetist proceeded with spinal anaesthesia and subsequently the patient was assessed in the recovery ward for evidence of lumbar plexus blockade after resolution of spinal anaesthesia.

Either a 25 G Whitacre or a 22 G Yale needle was used to administer 17.5 mg bupivacaine 0.5% intrathecally. Rectal paracetamol 1 g and diclofenac 75 mg i.v. were then administered.

Intraoperative hypotension was defined as a decrease in systolic blood pressure from a baseline of ≥30% and was treated with increments of ephedrine 6 mg i.v. Further midazolam 1–2 mg i.v. was titrated to a maximum total dose of 10 mg as clinically indicated. Maintenance and replacement fluids consisted of Hartmann's solution and were administered at the discretion of the attending anaesthetist.

After surgery the following data were recorded 8, 14, 20 and 24 h after performance of the PCB, by nursing staff unaware of the group allocations: time to first morphine analgesia, femoral nerve and lateral femoral cutaneous (LFC) nerve sensory block, obturator nerve motor block, pain scores at rest, presence of nausea/vomiting, sedation, heart rate, blood pressure and 24-h cumulative morphine consumption. Patients were cared for in a five-bed recovery ward for the first 24 h after surgery.

Sensory block of the femoral nerve (anterior thigh) and LFC nerve (lateral thigh) were evaluated using ethyl chloride spray and compared with the contralateral leg. Sensation was scored as described earlier. Blockade of the obturator nerve was evaluated by degree of motor block. Patients were asked to adduct their thigh and scored as 0=normal power, 1=weakness in adduction, 2=paralysis of adduction. Offset of sensory block in the femoral and LFC nerve distributions was defined as return to normal sensation. Offset of obturator motor block was defined as return to normal adductor power. For the purpose of data analysis, persistence of sensory or motor block at 24 h was recorded as offset at 24 h.

Pain was assessed using a standard verbal rating score for all patients, with 0=no pain and 10=worst pain imaginable. Pain was assessed at rest only to avoid prosthesis dislocation. Nausea and/or vomiting were recorded as present or absent. Antiemetic administration (intramuscular cyclizine 50 mg every 8 h if required) in the 24-h period was recorded. Sedation was scored as 0=awake, 1=drowsy, 2=asleep but rousable and 3=comatose. No sedatives were administered in the first 24 h after block insertion. Hypotension was defined as described earlier. Bradycardia was defined as a heart rate less than 60 beats min–1.

Nursing staff administered paracetamol 1 g every 6 h and diclofenac 75 mg every 12 h irrespective of pain status. Patients were given morphine 0.15 mg kg–1 i.m. when they requested supplementary analgesia and at 4-hourly intervals thereafter as required. Duration of analgesia was defined as the time from completion of block injection (time zero) to first administration of morphine. For the purpose of data analysis, patients who did not require morphine within 24 h had the time to first morphine analgesia recorded as 24 h. Twenty-four hours after surgery, cumulative morphine consumption was noted.

The study sample size was determined from previous work10 and our own data.11 Based on {alpha}=0.05, ß=0.2 and seeking a difference of 3.5 h in the interval to first morphine administration (representing an increase of ~40% in PCB analgesic duration) with an estimated standard deviation of 4 h, a sample size of 19 per group was calculated.

Statistical analysis was performed with GraphPad Prism version 4.00 for Windows (GraphPad Software, San Diego, CA USA; www.graphpad.com). Continuous parametric data were analysed using one-way ANOVA with Tukey's test for post-testing. Non-parametric data were analysed using the Kruskal–Wallis test with Dunn's test for post-testing. Proportional data were analysed using Fisher's exact test. Data are presented as mean (SD) or median (range). P<0.05 was taken as statistically significant.


    Results
 Top
 Footnotes
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
After randomization, 21 patients were allocated to group L, 20 patients to group T and 19 patients to group IT. The three groups were similar in terms of age, gender, weight, ASA classification, and surgery type (Table 1). Lumbar plexus block and spinal anaesthesia were successful in all patients.


View this table:
[in this window]
[in a new window]
 
Table 1 Patient characteristics. Data are mean (range) or mean (SD). No significant differences

 
The interval [h; mean (SD)] from time of completion of block injection to first morphine administration was similar in group L [11.2 (6.6)], group T [14.5 (8.0)] and group IT [14.6 (6.8)] (P=0.35). Twenty-four-hour cumulative morphine (mg) requirements [mean (SD)] were also similar between the three groups [group L, 21.9 (10.1); group T, 19.8 (6.7); group IT, 16.5 (9.5)]. These data are only for those patients who received morphine (group L, n=18; group T, n=12; group IT, n=15). The number of patients not receiving morphine within 24 h was seven in group T, four in group IT and three in group L (P=0.28). Onset of femoral nerve sensory block occurred 7.5 min after block in 11 patients in group L, six patients in group T (P=0.21) and eight patients in group IT (P=0.55). The duration of femoral or LFC nerve sensory block or obturator nerve motor block was similar between groups (Table 2).


View this table:
[in this window]
[in a new window]
 
Table 2 Duration (h) of sensory and motor block. Data are median (range). No significant differences

 
There were no differences in pain scores between groups at any of the four time points except at 20 h (group T vs group L, P=0.048) and 24 h (group T vs group IT, P=0.047) after block (Table 3).


View this table:
[in this window]
[in a new window]
 
Table 3 Pain verbal rating scores. Data are median (range). T=time (h) after block.

 
Sedation scores [median (range)] were 0 (0–2) for all groups at each time point, except at 14 h, when patients in group IT [1 (0–2)] were more sedated than those in group L [0 (0–2)] (P=0.02).

Adverse effects, including episodes of intraoperative hypotension, indicated by ephedrine requirements were also similar in terms of nature and incidence in the three groups (Table 4).


View this table:
[in this window]
[in a new window]
 
Table 4 Ephedrine usage and number of patients with adverse effects (n). Data are mean (SD). ; ;

 

    Discussion
 Top
 Footnotes
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
The results of this study indicate that the use of tramadol as a perineural or i.v. adjunct neither prolongs nor augments analgesia or anaesthesia in the setting of PCB with levobupivacaine 0.5%.

There are a number of possible mechanisms of action supporting the rationale for using tramadol as an adjunct for regional anaesthesia. First, tramadol's monoaminergic actions include agonism at peripheral {alpha}2 receptors, suggesting a role in nerve blocks similar to that of clonidine.13 Secondly, the presence of serotonin (5-hydroxytryptamine, 5-HT) subtype 3 (5-HT3) receptors on peripheral nerve endings and in the dorsal laminae of the spinal cord indicates possible peripheral sites of analgesic action for tramadol. Tramadol increases 5-HT concentrations both by inhibiting reuptake and stimulating release. The role of 5-HT3 receptors in analgesia is further supported by the findings that the 5-HT3 antagonist ondansetron inhibits tramadol's analgesic effects.14 Finally, tramadol has local anaesthetic properties possibly by blocking K+ channels.15 The direct application of tramadol to rat sciatic nerves results in conduction block,6 and in humans intradermal administration of tramadol provides local anaesthesia for minor skin procedures.16

Direct comparison of studies involving tramadol for regional anaesthesia is difficult, as duration of nerve block has been defined in terms of duration of postoperative analgesia9 10 or of duration of anaesthesia.7 Furthermore, few studies have included a systemic tramadol group.7 17

Studies of adjunctive tramadol in peripheral nerve blocks are few and have investigated tramadol's effects when combined with a local anaesthetic. Kapral and colleagues demonstrated that the addition of tramadol 100 mg to mepivacaine 1% for axillary plexus block prolonged sensory and motor block compared with mepivacaine alone or axillary block with tramadol 100 mg i.v.7 However, Robaux and colleagues failed to demonstrate an increase in sensory or motor block when tramadol was combined with mepivacaine 1.5% for brachial plexus block. Tramadol increased postoperative analgesia duration in a dose-dependent manner; however, a systemic tramadol group was not included in their study.8

Our finding that tramadol did not prolong analgesia or anaesthesia in the setting of PCB is further supported by the fact that no differences in block onset, 24-h morphine consumption or pain scores between the tramadol and control groups were found. Application of our findings in other clinical settings may not be appropriate. The local anaesthetic we used, levobupivacaine, is much longer-acting than mepivacaine and also has vasoconstrictor properties.18 The long duration of analgesia (11 h), as measured by time to first morphine administration, provided by levobupivacaine alone may have obscured a possible effect of tramadol. Tramadol added to ropivacaine 0.75% for brachial plexus block provided only ~9 h of analgesia.4 Secondly, the administration of tramadol into a compartment compared with placement in the brachial plexus sheath may affect drug concentrations. The concentration of tramadol 0.375% used in our study is less than the concentration of 0.5% that was the most effective dose in the study of Robaux and colleagues.8 Direct application of tramadol 1.25% in contrast to 2.5 and 5% to rat sciatic nerve failed to demonstrate a difference in amplitude reduction compared with saline.6 The physical dispersal of solution within the psoas compartment may affect local neural concentrations with consequent reductions in its direct neural effects.11 Our use of tramadol 1.5 mg kg–1 was based on dosages described in the literature at the commencement of our study.

We chose to study patients undergoing hip and knee arthroplasty as PCB provides excellent postoperative analgesia after both procedures with median visual analogue scores (VAS) of 10 mm at 24 h.12 19 While both joints receive innervation from the lumbar plexus and sciatic nerve,2022 studies show that blockade of the sciatic component is clinically unimportant for postoperative analgesia.12 22 However, the addition of obturator nerve blockade to femoral nerve block provides superior analgesia than femoral block alone.12 23 We believe that in the setting of peripheral nerve blockade the efficacy of postoperative analgesia is comparable for hip and knee arthroplasty.

The power of our study was based on an original estimate of a standard deviation of 4 h for time to first morphine analgesia. The standard deviation varied from 6.6 to 8 h; therefore it is possible that a type II error may have occurred. We had estimated the standard deviation from preliminary data investigating clonidine's effect on PCB and from work investigating tramadol combined with caudal bupivacaine.10 11 For the study to have a power of 80%, 63 patients would have been required per group.

The measurement of pain using VAS is common but has proved difficult to administer in the older orthopaedic population, with over half of patients unable to complete a VAS.24 A verbal numerical scale was used as these have been shown to correlate with VAS in a number of patient settings,2527 and we have used this scale successfully in a similar setting.11

In conclusion, our data do not support a clinically important local anaesthetic or peripheral analgesic effect of tramadol as adjunct to psoas compartment block with levobupivacaine 0.5%.


    Acknowledgments
 
We would like to thank the nursing staff of Block 9, St Mary's Orthopaedic Hospital, for their time and assistance with data collection.


    Footnotes
 Top
 Footnotes
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
{dagger} Scientific Presentations—Presented at the XXIII ESRA Annual Congress, Athens, 8–11 September 2004. Back

{ddagger} Declaration of interest. The Department of Anaesthesia at University College Cork has received educational grants from Abbott Laboratories, Ireland. Back


    References
 Top
 Footnotes
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
1 Murphy DB, McCartney CJ, Chan VW. Novel analgesic adjuncts for brachial plexus block: a systematic review. Anesth Analg 2000; 90: 1122–8[Abstract/Free Full Text]

2 Erlacher W, Schuschnig C, Koinig H, et al. Clonidine as adjuvant for mepivacaine, ropivacaine and bupivacaine in axillary, perivascular brachial plexus block. Can J Anaesth 2001; 48: 522–5[Abstract/Free Full Text]

3 Culebras X, Van Gessel E, Hoffmeyer P, Gamulin Z. Clonidine combined with a long acting local anesthetic does not prolong postoperative analgesia after brachial plexus block but does induce hemodynamic changes. Anesth Analg 2001; 92: 199–204[Abstract/Free Full Text]

4 Antonucci S. Adjuvants in the axillary brachial plexus blockade. Comparison between clonidine, sufentanil and tramadol. Minerva Anestesiol 2001; 67: 23–7[Medline]

5 Budd K, Langford R. Tramadol revisited. Br J Anaesth 1999; 82: 493–5[Free Full Text]

6 Tsai YC, Chang PJ, Jou IM. Direct tramadol application on sciatic nerve inhibits spinal somatosensory evoked potentials in rats. Anesth Analg 2001; 92: 1547–51[Abstract/Free Full Text]

7 Kapral S, Gollmann G, Waltl B, et al. Tramadol added to mepivacaine prolongs the duration of an axillary brachial plexus blockade. Anesth Analg 1999; 88: 853–6[Abstract/Free Full Text]

8 Robaux S, Blunt C, Viel E, et al. Tramadol added to 1.5% mepivacaine for axillary brachial plexus block improves postoperative analgesia dose-dependently. Anesth Analg 2004; 98: 1172–7[Abstract/Free Full Text]

9 Alhashemi JA, Kaki AM. Effect of intrathecal tramadol administration on postoperative pain after transurethral resection of prostate. Br J Anaesth 2003; 91: 536–40[Abstract/Free Full Text]

10 Senel AC, Akyol A, Dohman D, Solak M. Caudal bupivacaine-tramadol combination for postoperative analgesia in pediatric herniorrhaphy. Acta Anaesthesiol Scand 2001; 45: 786–9[CrossRef][ISI][Medline]

11 Mannion S, Hayes I, Loughnane F, Murphy DB, Shorten GD. Intravenous but not perineural clonidine prolongs postoperative analgesia after psoas compartment block with 0.5% levobupivacaine for hip fracture surgery. Anesth Analg 2005 (In press)

12 Capdevila X, Macaire P, Dadure C, et al. Continuous psoas compartment block for postoperative analgesia after total hip arthroplasty: new landmarks, technical guidelines, and clinical evaluation. Anesth Analg 2002; 94: 1606–13[Abstract/Free Full Text]

13 Kayser V, Besson JM, Guilbaud G. Evidence for noradrenergic component in the antinociceptive effect of the analgesic agent tramadol in an animal model of clinical pain. Eur J Pharmacol 1992; 224: 83–8[CrossRef][ISI][Medline]

14 Arcioni R, della Rocca M, Romano S, Romano R, Pietropaoli P, Gasparetto A. Ondansetron inhibits the analgesic effects of tramadol: a possible 5-HT3 spinal receptor involvement in acute pain in humans. Anesth Analg 2002; 94: 1553–7[Abstract/Free Full Text]

15 Mert T, Gunes Y, Guven M, Gunay I, Gocmen C. Differential effects of lidocaine and tramadol on modified nerve impulse by 4-aminopyridine in rats. Pharmacology 2003; 69: 68–73[CrossRef][ISI][Medline]

16 Altunkaya H, Ozer Y, Kargi E, Babuccu O. Comparison of local anaesthetic effects of tramadol with prilocaine for minor surgical procedures. Br J Anaesth 2003; 90: 320–2[Abstract/Free Full Text]

17 Gunes Y, Gunduz M, Unlugenc H, Ozalevli M, Ozcengiz D. Comparison of caudal vs intravenous tramadol administered either preoperatively or postoperatively for pain relief in boys. Paediatr Anaesth 2004; 14: 324–328[CrossRef][Medline]

18 Foster RH, Markham A. Levobupivacaine: a review of its pharmacology and use as a local anaesthetic. Drugs 2000; 59: 551–79[ISI][Medline]

19 Kaloul I, Guay J, Cote C, Fallaha M. The posterior lumbar plexus (psoas compartment) block and the three-in-one femoral nerve block provide similar postoperative analgesia after total knee replacement. Can J Anaesth 2004; 51: 45–51[Abstract/Free Full Text]

20 Birnbaum K, Prescher A, Hessler S, Heller KD. The sensory innervation of the hip joint: an anatomical study. Surg Radiol Anat 1997; 19: 371–5[ISI][Medline]

21 Pick TP, Howden R, eds. Gray's Anatomy. Leicester: Galley Press, 1988

22 Allen HW, Liu SS,Ware PD, Nairn CS, Owens BD. Peripheral nerve blocks improve analgesia after total knee replacement surgery. Anesth Analg 1998; 87: 93–7[Abstract]

23 Macalou D, Trueck S, Meuret P, et al. Postoperative analgesia after total knee replacement: the effect of an obturator nerve block added to the femoral 3-in-1 nerve block. Anesth Analg 2004; 99: 251–4[Abstract/Free Full Text]

24 de Visme V, Picart F, Le Jouan R, Legrand A, Savry C, Morin V. Combined lumbar and sacral plexus block compared with plain bupivacaine spinal anesthesia for hip fractures in the elderly. Reg Anesth Pain Med 2000; 25: 158–62[CrossRef][ISI][Medline]

25 De Conno F, Caraceni A, Gamba A, et al. Pain measurement in cancer patients: a comparison of six methods. Pain 1994; 57: 161–6[CrossRef][ISI][Medline]

26 Kunst G, Chrubasik S, Black AM, Chrubasik J, Schulte-Monting J, Alexander JI. Patient-controlled epidural diamorphine for post-operative pain: verbal rating and visual analogue assessments of pain. Eur J Anaesthesiol 1996; 13: 117–29[ISI][Medline]

27 Holdgate A, Asha S, Craig J, Thompson J. Comparison of a verbal numeric rating scale with the visual analogue scale for the measurement of acute pain. Emerg Med 2003; 15: 441–6





This Article
Abstract
Full Text (PDF)
All Versions of this Article:
94/3/352    most recent
aei057v1
E-Letters: Submit a response to the article
Alert me when this article is cited
Alert me when E-letters are posted
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Disclaimer
Request Permissions
Google Scholar
Articles by Mannion, S.
Articles by Shorten, G. D.
PubMed
PubMed Citation
Articles by Mannion, S.
Articles by Shorten, G. D.