Article Text

Download PDFPDF

Intercostal chest drain insertion by general physicians: attitudes, experience and implications for training, service and patient safety
  1. John P Corcoran1,
  2. Robert J Hallifax2,
  3. Ambika Talwar3,
  4. Ioannis Psallidas1,
  5. Annemarie Sykes1,
  6. Najib M Rahman1
  1. 1Oxford Centre for Respiratory Medicine, Oxford University Hospitals NHS Trust, Oxford, UK
  2. 2Department of Respiratory Medicine, Buckinghamshire Healthcare NHS Trust, Aylesbury, UK
  3. 3Department of Respiratory Medicine, Royal Berkshire NHS Foundation Trust, Reading, UK
  1. Correspondence to Dr John P Corcoran, Oxford Centre for Respiratory Medicine, Churchill Hospital, Oxford OX3 7LE, UK; jpcorcoran{at}


Background Intercostal chest drain (ICD) insertion is considered a core skill for the general physician. Recent guidelines have highlighted the risks of this procedure, while UK medical trainees have reported a concurrent decline in training opportunities and confidence in their procedural skills.

Objectives We explored clinicians’ attitudes, experience and knowledge relating to pleural interventions and ICD insertion in order to determine what changes might be needed to maintain patient safety and quality of training.

Methods Consultants and trainees delivering general medical services across five hospitals in England were invited to complete a questionnaire survey over a 5-week period in July and August 2014.

Results 117 general physicians (32.4% of potential participants; comprising 31 consultants, 48 higher specialty trainees, 38 core trainees) responded. Respondents of all grades regarded ICD insertion as a core procedural skill. Respondents were asked to set a minimum requirement for achieving and maintaining independence at ICD insertion; however, only 25% of higher specialty trainees reported being able to attain this self-imposed standard. A knowledge gap was also revealed, with trainees managing clinical scenarios correctly in only 51% of cases.

Conclusions Given the disparity between clinical reality and what is expected of the physician-in-training, it is unclear whether ICD insertion can remain a core procedural skill for general physicians. Consideration should be given to how healthcare providers and training programmes might address issues relating to clinical experience and knowledge given the implications for patient safety and service provision.


Statistics from


Pleural disease is common, affecting >3000 people per million population annually. The number of patients with symptomatic pleural disease requiring intervention is increasing, and this trend is expected to continue. The incidence of pleural infection in adult and paediatric populations has doubled over the past decade.1 A year-on-year increase in new cancer diagnoses, better long-term survival with treatment and frequent involvement of the pleura in metastatic disease2–4 means >40 000 people in the UK annually will develop a malignant effusion. The majority of these patients will be symptomatic, and the range of treatment options means specialist input is often needed to decide what intervention is most appropriate.5

The end result is an increasing number of patients with pleural disease presenting to general physicians via the acute take. As such, the ability to manage conditions such as pleural effusion and pneumothorax remains a key component of core and higher specialty medical training curricula in the UK, including that for general internal medicine (GIM).6 This includes learning procedural skills such as pleural aspiration and intercostal chest drain (ICD) insertion. However, trainees have reported increasing difficulty accessing procedural training opportunities with a corresponding decline in confidence when performing these interventions.7–9

Alongside these training issues is the recognition that pleural procedures are associated with potentially serious harm to patients (table 1). A National Patient Safety Agency (NPSA) report noted widespread issues with suboptimal ICD management by inexperienced ward staff, as well as 27 cases of serious harm or death related to procedural complications over a 3-year period.10 A subsequent survey raised further concerns regarding consent, training and major iatrogenic complications related to ICD insertion in the majority of UK hospitals.11 The vast majority of complications from pleural procedures are easily avoidable and frequently caused through inexperience, poor technique and/or inappropriate intervention.12 Recent work from the Mayo Clinic in the USA showed a significant reduction in iatrogenic complications could be achieved by simply limiting the number of physicians authorised to perform pleural procedures.13 This realisation has prompted measures to improve patient safety, notably the recommendation in recent guidelines that thoracic ultrasound (TUS) should be used during all pleural interventions for suspected fluid.14 This has been supported by evidence showing TUS reduces the risk of complications from pleural procedures,15 ,16 leading commentators to question whether failure to use TUS is medicolegally defensible in this clinical setting.17

Table 1

Examples of common procedural complications associated with intercostal chest drain (ICD) insertion and preventative strategies

The requirement for TUS during pleural procedures has predictably affected trainees’ experience. The current UK training curriculum for GIM6 does not consider TUS when stipulating the need for trainees to develop competence at pleural interventions. Other than those accrediting in respiratory medicine, the majority of GIM trainees are not required or expected to be competent at using TUS. This inevitably limits training opportunities and experience, lowering the confidence of trainees already concerned about potential complications. This has led to increased input from specialist teams with the necessary skill set (including TUS) to maintain service provision,18–20 further reducing the exposure of trainees to these procedures and creating a cycle of declining experience, confidence and knowledge.

In the context of these changes in clinical practice, we explored clinicians’ attitudes, experience and knowledge relating to pleural interventions and specifically ICD insertion. The objectives of this work were to quantify trainees’ procedural experience and expectations in order to determine what changes might be needed to improve training, patient safety and service delivery.



We approached consultants and trainees (core medical and higher specialty including a GIM training component) working in acute general medical services across five hospitals in southern England over a 5-week period in July and August 2014. Potential participants were identified at GIM-related educational or service activities including medical grand rounds, regional GIM training days, clinical governance meetings and clinical handover meetings. Questionnaires were distributed by hand and electronically to those individuals who were eligible to participate; non-responders were not followed up.


A questionnaire survey was developed and piloted in a single location (Oxford University Hospitals National Health Service (NHS) Trust) before amendment and wider distribution. Demographic data collected from participants included their field of specialty interest and year of training (trainees) or being awarded their Certificate of Completion of Training (CCT) (consultants).


All participants were asked a series of questions assessing their attitudes towards ICD insertion and whether they considered it an integral and/or valuable skill for a GIM trainee. Answers were provided using a five-point Likert-type scale ranging from 1 (strongly agree) to 5 (strongly disagree).

Training and experience

All participants were asked how many procedures were necessary to attain and subsequently maintain independent competence in ICD insertion. Trainees only were asked additional questions relating to their experience and confidence with ICD insertion and TUS.

Clinical knowledge

Five case scenarios of patients with conditions commonly seen in an acute admissions unit (pleural effusion, pleural infection and pneumothorax) were presented. Trainees only were asked to choose the best of five prespecified answers based on their interpretation of the clinical information provided. The correct answers were determined using British Thoracic Society (BTS) guidelines5 ,21 ,22 and independently verified by three respiratory physicians with a subspecialty interest in pleural disease. None of the case scenarios presented was the subject of disagreement between either the guidelines or clinicians verifying the answers.

Data analysis

Data from the questionnaires were entered onto a spreadsheet and analysed using SPSS V.19.0. Tests of significance including t test and analysis of variance were used with a p value of <0.05 defined as significant. Responses from all participants were analysed to determine clinicians’ attitudes towards ICD insertion, expectation of GIM trainees and criteria for procedural competency. Responses from trainees were additionally analysed to assess levels of procedural experience and confidence. Responses to the case scenarios were analysed to determine trainees’ knowledge of pleural disease and its management.


Participants and demographics

A total of 117/361 clinicians responded to the survey, representing 32.8% of eligible participants across the five hospitals. Thirty-one consultants with a CCT in GIM replied, of whom 8/31 (25.8%) had a CCT in respiratory medicine and 23/31 (74.2%) a CCT in other medical specialties. Also, 19/31 (61.3%) consultants had >5 years’ experience working at this level and finished their training before the publication of the NPSA report relating to ICDs.7 Forty-eight higher specialty trainees (HSTs—ie, ST3 level or above) replied, of whom 10/48 (20.8%) were dual-accrediting in respiratory medicine and GIM; 38/48 (79.2%) were non-respiratory HSTs accrediting in GIM. 20/48 (41.7%) HSTs were ST3/4 trainees, with 28/48 (58.3%) being ST5 level or above. Thirty-eight core medical trainees (CMTs) replied, with 20/38 (52.6%) of these being CT1 and 18/38 (47.4%) CT2s.


Clinicians of all grades agreed ICD insertion was a core procedural skill that all registrars should be independently competent in prior to being granted a CCT in GIM; consultants agreed with this more strongly than trainees (1.74±1.26 vs 2.31±1.14 (mean±SD), p=0.02). No significant difference was observed between responses according to field of specialty interest across consultants and HSTs. However, clinicians of all grades were uncertain when asked whether ICD insertion should become a specialist skill performed by a limited number of doctors (table 2).

Table 2

Attitudes towards intercostal chest drain (ICD) insertion among survey respondents according to grade

Training and experience

The majority of clinicians felt a doctor would need to have inserted between 5 and 10 ICDs before achieving independent competence; subsequently, between 5 and 10 ICDs would need to be inserted on an annual basis to maintain that competence (table 3). When questioned about their current training, 43/48 (89.6%) of GIM-accrediting HSTs stated they had enough baseline experience (ie, inserted five or more ICDs during their training) while 12/48 (25.0%) had inserted enough ICDs in the previous 12 months to maintain their independent competence. Of the 12/48 (25.0%) HSTs who achieved this latter standard, 10/12 (83.3%) were dual-accrediting in respiratory medicine and GIM. Only 2/38 (5.3%) non-respiratory HSTs had been able to maintain their competence at ICD insertion according to the basic standard proposed by clinicians responding to this survey. Overall, HSTs remained largely confident in their ability to insert an ICD in an emergency for either a pleural effusion or pneumothorax (table 4).

Table 3

Questions relating to training competency in ICD insertion (all clinicians)

Table 4

Questions relating to current training and experience in ICD insertion (trainees only)

Almost half of the trainees (41/86, 47.7%) reported having never had access to simulation training in ICD insertion. Nine HSTs—eight of whom were accrediting in respiratory medicine—had completed formal training in the use of TUS. The majority of trainee respondents (49/86, 57.0%) had never received any TUS training and did not expect to in the future (table 4).

Clinical knowledge

All but one of the trainee respondents completed the clinical questionnaire (table 5), the results of which are presented in figure 1. In the management of pneumothorax, only 44/170 (25.9%) answers across these two cases were in line with the best response available. In total, 78/170 (45.9%) of responses favoured the unnecessary use of ultrasound guidance during a therapeutic intervention for pneumothorax. In treatment choices for cases involving pleural effusions, 115/170 (67.6%) of answers were the best response available; in all three of these clinical scenarios, CMTs marginally outperformed HSTs.

Table 5

Clinical questionnaire relating to pleural disease and interventions using best-of-five format (trainees only)

Figure 1

Trainee responses to best-of-five clinical questionnaire (see table 5) relating to pleural disease and interventions. CMT, core medical trainee; HST, higher specialty trainee; ICD, intercostal chest drain; US, ultrasound.


Our study shows the majority of current GIM trainees are not obtaining sufficient experience of ICD insertion to maintain competence according to a self-imposed standard. This contrasts with the apparent importance placed on this skill. Being able to perform ICD insertion continues to be associated by GIM physicians with competence to fulfil the responsibilities of being a registrar and progression to consultant level. However, there is a suggestion that physicians are ambivalent towards the idea of ICD insertion becoming a specialist intervention. This perhaps recognises a continuing decline in trainees’ experience, alongside recent changes in clinical practice that have occurred following the NPSA report into ICD insertion10 and national guideline updates.14

A move towards competency-based training over the past decade has probably influenced the attitude of consultants and trainees towards procedural work. The adoption of workplace-based assessments23 has focused trainees’ minds on the need to demonstrate continued clinical competence on a regular basis. At the same time, there is a desire for postgraduate medical training to be shorter and more general in nature. This is best encapsulated in the recent Shape of Training report.24 Such a move would further impact on procedural exposure and experience among trainees. Challenges already exist in developing and maintaining specialty-specific skills for dual-accrediting trainee physicians in the current system.25 ,26 ICD insertion becoming a specialist-only skill may be the result of junior doctors deciding their training priorities lie elsewhere and, for procedural skills, within their own specialties.

The results of this and other surveys7 ,8 have major implications for medical training. A consistent lack of exposure to ICD insertion raises the question of whether it is a skill all GIM trainees can realistically maintain. Outside specialties such as respiratory, emergency and intensive care medicine, trainees are unlikely to carry out ICD insertion frequently. Furthermore, the need to be competent in TUS to intervene in any case of suspected pleural fluid14 limits who can perform this procedure. Most trainees have little or no training in TUS, a situation that is unlikely to change unless it can be incorporated into medical undergraduate curricula.27 ,28 The increasing provision of specialist pleural teams, including nurse-led services in many UK hospitals,18–20 will further reduce training opportunities for junior doctors. Applying principles that have helped rationalise how key services, notably surgical, are delivered within the NHS, it may be that ICD insertion is best performed by those who do so regularly given the risks involved.10 ,11 ,13

Nonetheless, any restriction on the number of clinicians who can perform ICD insertion will have resource implications, particularly in smaller hospitals. As such, service demands are likely to dictate an ongoing role for GIM trainees in performing ICD insertion for the foreseeable future—particularly in the setting of an acutely compromised patient, for example, tension pneumothorax. Accepting this fact, consideration needs to be given as to how a large number of trainees can safely maintain their skills in the absence of regular patient exposure. Simulation training can improve confidence at invasive procedures including ICD insertion,29 ,30 although this does not necessarily correlate with real-world capability. Bespoke assessment tools31 ,32 may help in the assessment of competence, although these are yet to be validated on a wider scale. Furthermore, this and other surveys33 indicate access to simulation training facilities for pleural interventions may be limited. Providing this capacity for all GIM trainees nationally is likely to be difficult in the face of other training and service pressures, but needs addressing given the implications for service delivery and patient safety otherwise.

Practical competence in performing a procedure must be accompanied by an understanding of when it is indicated. Pleural disease is recognised as a subspecialty area within respiratory medicine and has moved away from being a generalist's field. This is reflected in deficiencies in GIM trainees’ knowledge, with an incorrect choice of management strategy seen frequently in the patient case scenarios presented in our questionnaire. This knowledge gap is probably linked to the limited clinical experience reported, and certainly reflected in a recent BTS pleural audit.34 This demonstrated similar findings to our survey and illustrated the potential for poor decision making to negatively impact on patient care with unnecessary intervention, exposure to risk and/or hospital admission. Any changes to medical training programmes must address clinical understanding as well as practical skills if patient safety is to be maintained.

There are limitations to this work. The number of clinicians who responded to the survey was moderate and all work in a limited number of hospitals. However, they are based in secondary and tertiary centres of varying sizes and the experience of trainees in particular is likely to be comparable with their counterparts across the UK. The problems identified with regard to procedural exposure, experience and knowledge are likely to be replicated elsewhere and merit consideration when assessing training standards and curricula. A national survey addressing exposure to ICD insertion and other invasive procedures among medical trainees may further inform individuals involved in maintaining training standards and patient safety.

The primary aim of physicians should be primum non nocere. Avoiding patient harm from ICD insertion means considering multiple factors. Service delivery requires a supply of clinicians who can safely perform ICD insertion in elective and emergency scenarios. Until now, this has been the GIM registrar—however, our survey suggests this cohort of doctors no longer has the experience necessary to fulfil this role. Hospitals therefore need to consider alternative means by which they can safely manage patients with pleural disease. Given the need for ultrasound competence to perform pleural interventions, these may include developing radiology services or specialist pleural teams with rapid access clinics and procedural lists. Regardless of the approach adopted, early specialist input should be encouraged since it can reduce harm and improve patient care. A restriction in the number of clinicians required to maintain competence in ICD insertion may limit the impact of any reduction in training opportunities. The development of simulation training and assessment models should allow further experience in a low-risk environment. Service delivery, patient safety and training are inexorably intertwined. This survey highlights the need for further debate as to how these varying needs are met for all pleural procedures including ICD insertion.

Main messages

  • Intercostal chest drain (ICD) insertion is considered a core procedural skill for general physicians.

  • Most trainees have difficulty accessing training opportunities in ICD insertion and cannot maintain procedural competence.

  • Most trainees have variable understanding about how to manage patients with common pleural conditions.

  • It is uncertain whether all general physicians can maintain competence in ICD insertion, and how this will impact on training, service and patient safety.

Current research questions

  • Can training programmes be modified to offer trainees adequate exposure to ICD insertion?

  • How can simulation models be developed to support training and maintenance of skills in ICD insertion?

  • Who should replace general physicians in performing ICD insertion and what are the implications for service and patient safety?

Key references

  • Wrightson JM, Fysh E, Maskell NA, et al. Risk reduction in pleural procedures: sonography, simulation and supervision. Curr Opin Pulm Med 2010;16:340–50

  • Havelock T, Teoh R, Laws D, et al., BTS Pleural Disease Guideline Group. Pleural procedures and thoracic ultrasound: British Thoracic Society Pleural Disease Guideline 2010. Thorax 2010;65(Suppl 2):ii61–76.

  • Hutton IA, Kenealy H, Wong C. Using simulation models to teach junior doctors how to insert chest tubes: a brief and effective training module. Intern Med J 2008;38:887–91.

  • Salamonsen MR, Bashirzadeh F, Ritchie AJ, et al. A new instrument to assess physician skill at chest tube insertion: the TUBE-iCOMPT. Thorax 2015;70:186–8.



  • Contributors JPC, AS and NMR conceived the project and drafted the initial manuscript. JPC, RJH, IP and AT performed data collection and analysis. All authors were involved in approving the final manuscript for submission. JPC and NMR are responsible for the overall content as guarantors.

  • Funding IP is the recipient of a European Respiratory Society Fellowship (LTRF 2013-1824). NMR is funded by the NIHR Oxford Biomedical Research Centre.

  • Competing interests None declared.

  • Ethics approval Formal ethics approval was not necessary as the study team were not required to access any confidential data, the questionnaire was anonymous and completion was taken to indicate consent to participate in the study.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.