Intercostal catheter insertion: are we really doing well?ans_6093 392..394

Rashid Alrahbi,* Ruth Easton,* Cino Bendinelli, Natalie Enninghorst, Krisztian Sisak and Zsolt J. BaloghDepartment of Traumatology, Division of Surgery, John Hunter Hospital and University of Newcastle, Newcastle, NSW, Australia

Key words

haemothorax, pneumothorax, thoracic injuries,thoracostomy, trauma.

Correspondence

Professor Zsolt J. Balogh, Department of Traumatology,Division of Surgery, John Hunter Hospital and Universityof Newcastle, Newcastle, NSW 2310, Australia. Email:zsolt.balogh@hnehealth.nsw.gov.au

R. Alrahbi MD; R. Easton BMed; C. Bendinelli MD,FRACS; N. Enninghorst MD; K. Sisak MD;Z. J. Balogh MD, PhD, FRACS.

This study was presented at the Royal AustralasianCollege of Surgeons Annual Scientific Congress inAdelaide, May 2011.

*These two authors contributed equally to this work.

Accepted for publication 4 March 2012.

doi: 10.1111/j.1445-2197.2012.06093.x

Abstract

Introduction: Intercostal catheters (ICC) are the standard management of chesttrauma, but are associated with complications in up to 30%. The aim of this study wasto evaluate errors in technique during ICC insertion to characterize the potentialbenefit of improved training programmes.Methods: Prospective audit of all ICC in trauma patients at a level 1 trauma centre forover 12 months. Exclusions were pigtail catheters and ICC inserted during thoracicsurgery. Errors were identified from patient examination and chest imaging; they weredefined as insertional, positional, incorrect size (<28 French) and lack of antibioticprophylaxis. Ongoing complications unrelated to an error in technique, for exampleblocked tube, were not analysed.Results: Fifty-seven patients received a total of 94 ICC during the study period.Patients were predominantly male (77%), mean age of 40 � 20 years, mean injuryseverity score 27 � 13, mean abbreviated injury scale chest 3.8 � 0.72. 86% wereblunt trauma and 14% penetrating chest injuries. Thirty-six errors in techniqueoccurred in 33 ICC insertions (38%). The most common errors were absence ofprophylactic antibiotics (13%), ICC too far out (9%), kinked (6%) and wrong-sizedICC (5%). Emergency had a significantly greater frequency of errors than otherspecialties (67%, relative risk 2.11, P = 0.002). The majority of ICC were inserted byregistrars, and registrars made a greater number of errors than fellows or consultants(relative risk 2.00, P = 0.02).Discussion: This study identified a large number of preventable errors for ICC inser-tion in trauma patients. Standardized institutional credentialing systems may berequired to ensure adequate proficiency of trainees performing this procedure.

Introduction

Intercostal catheter insertion (ICCI) is a life-saving procedure in themanagement of thoracic injuries, and is therefore a mandatory skillfor physicians involved in the care of injured patients. The frequencyof ICCI-related complications varies according to patient character-istics,1,2 location of insertion3,4 and the specialty of the operator,2,5–7

but has been reported to be as high as 25–30%.5,8,9

The UK National Patient Safety Agency (NPSA) released areport in 2008 in response to the high rate of complications asso-ciated with ICCI. After investigating case reports of mortality orserious harm, they highlighted concerns related to inadequate train-ing and experience of clinicians inserting drains, insufficient super-vision of junior staff, and a lack of adequate imaging andknowledge of clinical guidelines.10 Operators who perform ICCI inour institution are Emergency Management of Severe Trauma cer-tified but we do not provide further training or accreditation for

this invasive skill. We hypothesized that the doctors in our traumasystem perform ICCI with minimal errors. The aim of this studywas to evaluate the frequency and the nature of errors in techniqueduring ICCI, to identify the need for improvements in trainingprogrammes.

Methods

Ethics approval was obtained from the Hunter New EnglandHuman Research Ethics Committee before commencement of thestudy. A prospective audit of all ICCI in trauma patients at a Level1 trauma centre was performed over a 12-month period fromDecember 2009 to December 2010. All patients requiring traumateam activation with an ICCI before transfer or during admissionwere included. Exclusion criteria were pigtail catheters, andpatients with intercostal catheter (ICC) removed before arrival atthe trauma centre.

SPECIAL ARTICLEANZJSurg.com

© 2012 The AuthorsANZ Journal of Surgery © 2012 Royal Australasian College of SurgeonsANZ J Surg 82 (2012) 392–394

Data was prospectively collected on a daily basis, with patientdemographics, markers of injury severity (injury severity score(ISS), abbreviated injury scale (AIS) chest, vital signs, endotrachealintubation), hospital length of stay and ICU length of stay recorded.The time and indication for ICCI, the location where ICCI wasperformed, and the seniority and specialty of the operator were alsonoted. Errors were identified by patient examination and evaluationof pre- and post-insertion chest imaging.

The standard of care regarding appropriate ICCI was based on theBritish Thoracic Society guidelines.11 Preventable errors in tech-nique were therefore defined as follows:

• Insertional: ICC not inserted far enough (holes visible), inad-equate fixation, insertion through same incision as previousICC

• Positional: extrathoracic placement, wrong intercostal space

• Incorrect size of ICC (should be at least 28 French in adults)

• Absence of antibiotic prophylaxis.The definition mentioned earlier deliberately excludes complica-

tions that may occur even with perfect operator technique, forexample blocked tube, infection, etc. The sequelae associated withthese errors were not included, as the aim of the study was toevaluate compliance with the guidelines for ICCI, rather than tovalidate current recommendations.

All data analyses were performed using a commercial statisticalprogram (Stata IC 11, StataCorp, College Station, TX, USA; 2009).Categorical variables were analysed using relative risks and P valuescalculated using Fisher’s exact test. Continuous variables are pre-sented as mean � standard deviation, and were analysed using anunpaired t-test. Two-tailed P values were calculated. Statistical sig-nificance was determined at P < 0.05.

Results

Fifty-seven patients received a total of 94 ICC during the studyperiod. Patient characteristics are shown in Table 1. The patientpopulation was predominantly male (77%), with a mean age of 40 �

20 years and mean ISS of 27 � 13. Eighty-six percent of patientsexperienced blunt trauma and 14% penetrating chest injuries, with amean AIS chest of 3.8 � 0.72 standard devitaion. Forty-two patients

(73.7%) required unilateral ICCI, while bilateral ICCI insertion wasrequired in 15 patients (26.3%). Twenty-four patients required mul-tiple ICCI: three ICCI in six patients, four ICCI in two patients andfive ICCI in a further two patients.

The indications for ICCI are shown in Table 2, with the majorityinserted for pneumothorax and haemothorax. One patient presentedwith cardiac tamponade following a stab wound to the chest and wastaken to theatre for a sternotomy and open repair. Two ICC wereplaced by the cardiothoracic surgeon during the procedure. Anotherunusual case was a diaphragmatic rupture. This patient also under-went ICCI intraoperatively in conjunction with a laparotomy anddiaphragm repair.

Thirty-six errors in technique occurred in 33 ICC insertions(38% of total ICC). The breakdown of errors is shown in Table 3,with the most common being the absence of prophylactic antibiot-ics (13% of insertions), ICC too far out (9%), kinked tube (6%) anduse of the wrong-sized ICC (5%). There were no significant asso-ciation between errors and patient characteristics (Table 4). Table 5depicts the frequency of errors according operator professionallevel and specialty, and Table 6 shows errors according to the loca-tion and timing of ICC insertion. Emergency physicians and train-ees had a significantly greater frequency of errors than otherspecialties (67%, relative risk 2.11, P = 0.002), and likewise, therewas a greater frequency of errors for ICCI in the emergency depart-ment compared with other hospital locations (P = 0.05). The major-ity of ICC were inserted by registrars, and registrars made a greaternumber of errors than fellows or consultants (relative risk 2.00, P =0.02). A minority of ICC were inserted during working hours, andno significant association was seen between timing and rates oferror (Table 6).

Table 1 Patient characteristics

Study population (57 patients)

Male 44 (77%)Age (years) 39.5 � 19.9Injury severity score 26.9 � 13.2Hospital length of stay (days) 20 � 17.7Endotracheal intubation 32 (56%)Base deficit <-6 15 (26%)Intensive care admission 33 (58%)Intensive care length of stay (days) 10 � 11.3AIS chest 3.8 � 0.72Blunt chest injury 50 (86%)Penetrating chest injury 7 (14%)Unilateral ICC 42 (74%)Bilateral ICC 15 (26%)Multiple ICC 24 (42%)

AIS, abbreviated injury scale; ICC, intercostal catheter

Table 2 Indications for ICCI (n = 94)

Pneumothorax 46 (49%)Haemothorax 15 (16%)Haemopneumothorax 14 (15%)Tension pneumothorax 9 (10%)Empyema 4 (4.3%)Cardiac Tamponade 2 (2.1%)Haemodynamic instability 2 (2.1%)Massive surgical emphysema 1 (1.1%)Diaphragmatic rupture 1 (1.1%)

Table 3 Frequency of errors in technique

Error in technique Frequency (% of ICC)

No prophylactic antibiotics 12 (13)ICC too far out (hole visible) 8 (9)Kinked 6 (6)Wrong-sized ICC 5 (5)Extrathoracic (outside pleural cavity within

subcutaneous tissue)2 (2)

Intra-abdominal placement 1 (1)Inadequate fixation 1 (1)Insertion through previous ICC incision site 1 (1)Total 36 (38)

ICC, intercostal catheter.

Intercostal catheters: are we doing well? 393

© 2012 The AuthorsANZ Journal of Surgery © 2012 Royal Australasian College of Surgeons

Discussion

This study identifies a large number of preventable errors amongstaff performing ICCI in trauma patients. Emergency was shown tobe a specialty with a higher frequency of errors in technique, whichis consistent with previous literature.2,7 Our demonstration thatsenior staff made fewer errors substantiates the concerns of theNPSA that junior operators may often be inadequately trained andsupervised. ICCI is a fundamental skill in the management of chesttrauma; we suggest that standardized institutional credentialing

systems may be required to ensure adequate proficiency of traineesperforming this procedure. The high rate of errors also suggests thateven registrars already performing ICCI should be targeted forfurther training.

Limitations of the study include the absence of follow-up linkingerrors in technique to ongoing clinical sequelae including infectionand mortality. Several studies have already analysed the rate ofcomplications of ICCI and commented that it may be unacceptablyhigh. Our aim here was not to replicate their findings, but to furtherexplore the issue by characterizing the rate of clearly preventableerrors. In doing so, we could evaluate whether a simple retraining ofstaff would have the potential for significant improvement. Anotherlimitation of the study is the fact that errors were determined prima-rily through a review of the patient post-procedure, in combinationwith their medical record and imaging studies. Direct observation ofall procedures would likely have detected an even greater number oferrors, for example breach of the sterile field.

Overall, our data suggest that there are a large number of errorsduring ICCI that are avoidable, and identify a lack of staff training,which needs to be addressed. Senior hands-on input and supervisionis also advisable based on the fact that errors occur more frequentlyamong junior staff. At our institution, we are aiming to establish aprogramme based on simulated ICCI with a very realistic manne-quin (TraumaMan, Simulab Corporation, Seattle, WA, USA), andthen re-evaluate the rates of errors and complications. Our data andthe existing literature would suggest that emergency trainees maybenefit most and should be targeted in such an intervention.

References1. Heng K, Bystrzycki A, Fitzgerald M et al. Complications of intercostal

catheter insertion using EMST techniques for chest trauma. ANZ J. Surg.2004; 74: 420–3.

2. Etoch SW, Bar-Natan MF, Miller FB, Richardson JD. Tube thoracos-tomy. Factors related to complications. Arch. Surg. 1995; 130: 521–5.

3. Aylwin CJ, Brohi K, Davies GD, Walsh MS. Pre-hospital and in-hospitalthoracostomy: indications and complications. Ann. R. Coll. Surg. Engl.2008; 90: 54–7.

4. Chan L, Reilly KM, Henderson C, Kahn F, Salluzzo RF. Complicationrates of tube thoracostomy. Am. J. Emerg. Med. 1997; 15: 368–70.

5. Deneuville M. Morbidity of percutaneous tube thoracostomy in traumapatients. Eur. J. Cardiothorac Surg. 2002; 22: 673–8.

6. Maritz D, Wallis L, Hardcastle T. Complications of tube thoracostomyfor chest trauma. S. Afr. Med. J. 2009; 99: 114–7.

7. Ball CG, Lord J, Laupland KB et al. Chest tube complications: how wellare we training our residents? Can. J. Surg. 2007; 50: 450–8.

8. Helling TS, Gyles NR 3rd, Eisenstein CL, Soracco CA. Complicationsfollowing blunt and penetrating injuries in 216 victims of chest traumarequiring tube thoracostomy. J. Trauma 1989; 29: 1367–70.

9. Bailey R. Complications of tube thoracostomy in trauma. J. Accid.Emerg. Med. 2000; 17: 111–4.

10. National Patient Safety Agency. Rapid Response Report: Risks of ChestDrain Insertion. Reference NPSA/2008/RRR003, National PatientSafety Agency. 2008.

11. Laws D, Neville E, Duffy J, Pleural Diseases Group, Standards of CareCommittee, British Thoracic Society. BTS guidelines for the insertion ofa chest drain. Thorax 2003; 58: 53–9.

Table 4 Frequency of errors according to patient characteristics

Errors (36) No error (58) P value

Age 40 � 20 37 � 18 0.47Males 26 (72%) 47 (81%) 0.32Injury severity score 28 � 13 27 � 12 0.70AIS chest 3.8 � 0.9 3.8 � 0.8 1.00Intubated 22 (61%) 44 (76%) 0.17SBP <90 8 (24%) 15 (25%) 0.81Base deficit <-6 8 (22%) 17 (29%) 0.48Blunt injury 32 (89%) 52 (90%) 1.00Penetrating injury 4 (11%) 6 (10%)Pneumothorax 18 (50%) 28 (48%) 1.00Haemothorax 4 (11%) 11 (19%) 0.39Haemopneumothorax 4 (11%) 10 (17%) 0.55Tension pneumothorax 6 (17%) 3 (5%) 0.08Other 4 (11%) 6 (10%) 0.53

AIS, abbreviated injury scale; SBP, systolic blood pressure.

Table 5 Professional level and specialty of operators performing ICCI andrisk of errors

N(%)

Errors(%)

Relative risk(95%CI)

P value

OperatorSurgical 31 (33) 12 (39) 1.01 (0.59–1.75) 0.95Cardiothoracics 13 (14) 2 (15) 0.37 (0.10–1.35) 0.13Emergency 18 (19) 12 (67) 2.11 (1.33–3.36) 0.002Intensive care 12 (13) 5 (42) 1.10 (0.53–2.76) 0.79Trauma 20 (21) 5 (25) 0.60 (0.27–1.34) 0.21

All specialtiesRegistrar 50 (53) 25 (50) 2.00 (1.12–3.58) 0.02Fellow 15 (16) 4 (27) 0.66 (0.27–1.59) 0.35Consultant 29 (31) 7 (24) 0.54 (0.27–1.09) 0.09

CI, confidence interval; ICCI, intercostal catheter insertion.

Table 6 Location and timing of ICCI and risk of errors

N(%)

Errors(%)

Relative risk(95%CI)

P value

Emergency department 45 (48) 22 (49) 1.71 (1.00–2.92) 0.05Theatres 24 (25) 6 (25) 0.58 (0.28–1.23) 0.16Intensive care unit 18 (19) 5 (28) 0.68 (0.31–1.50) 0.34Wards 7 (7) 3 (43) 1.13 (0.46–2.77) 0.79Working hours 26 (28) 8 (31) 0.75 (0.39–1.42) 0.37Out of hours/weekends 68 (72) 28 (41) 1.34 (0.71–2.55)

CI, confidence interval; ICCI, intercostal catheter insertion.

394 Alrahbi et al.

© 2012 The AuthorsANZ Journal of Surgery © 2012 Royal Australasian College of Surgeons