A DESCRIPTIVE STUDY OF THE ACCURACY OF INTERCOSTAL DRAIN PLACEMENT AND FACTORS INFLUENCING PLACEMENT. Dr Tessa Korda, MBBCh 433192 Supervisors: Prof LN Goldstein Dr T Baillie Stanton A research report submitted to the Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, in partial fulfilment of the requirements for the degree of Master of Medicine in Emergency Medicine Johannesburg, September 2022 2 Declaration I, Dr Tessa Korda, hereby declare that this research report is my own work. It is being submitted for the degree of Master of Medicine in Emergency Medicine at the University of the Witwatersrand, Johannesburg. It has not been submitted before for any degree or examination at this or any other university. Dr Tessa Korda Signed on 16 September 2022 3 Contribution Declaration 4 Acknowledgements I would like to thank my supervisors, Prof. Goldstein and Dr Baillie Stanton, for their input and guidance throughout this process. Thank you to the emergency departments who allowed me to use the floor ultrasound and found space for me to work. To the doctors who so eagerly gave their time to participate – a big thank you. Thank you to the two simulated patients who were reliable and so easy to work with. To my partner, Bojan Korda, for his patience and support throughout. 5 Statement As per the University of the Witwatersrand Faculty of Health Sciences guidelines, this research report is being submitted in the following format: submission for publication ready format. The research report consists of a manuscript for submission, author’s guidelines, research protocol and annexures. The author’s guidelines and protocol are included for background reference. Manuscript for Submission Type of article: Original research Title of Manuscript: An observational simulation-based study describing the accuracy of intercostal drain placement and factors influencing placement. Word, Figure and Table Counts: Abstract: 272 words Manuscript: 2917 words Figures: 1 figure Tables: 2 tables 6 Table of Contents Declaration 2 Contribution declaration 3 Dedication 4 Statement 5 List of figures 7 List of tables 7 Abbreviations 7 Section 1: Manuscript for Submission 8 Background Methodology Results Discussion Conclusion References Section 2: Research Protocol 27 Appendix A Appendix B Annexures 41 A: HREC clearance certificate B: Turn-it-in plagiarism report C: Certificate of submission signed by student D: Study site permissions E: Separate copy of abstract F: Supervisor submission form G: ETD form 7 List of Figures Figure 1 - Photograph (A) and diagram (B) indicate how the right side of the thoracic wall was demarcated in the simulated patients List of Tables Table 1: Participant demographics Table 2: Accuracy of placement and placement in BTS triangle of safety Abbreviations AFJEM African Journal of Emergency Medicine ATLS Advanced Trauma Life Support BMI Body Mass Index BTS British Thoracic Society EC Emergency Centre ETC European Trauma Course ICD Intercostal Drain SA South Africa 8 Section 1 – Manuscript A descriptive study of the accuracy of intercostal drain placement and factors influencing placement. Tessa Korda, Tammy Baillie Stanton, Lara Goldstein Corresponding Author T Korda Division of Emergency Medicine School of Clinical Medicine Faculty of Health Sciences University of the Witwatersrand 7 York Road, Parktown Johannesburg 2193 Email: tessakorda@gmail.com Contact Number: +27 83 744 9203 9 Abstract Background South Africa’s significant burden of trauma and respiratory disease requires a high rate of intercostal drain (ICD) insertions by emergency centre practitioners. ICD misplacement is associated with significant complications. The aim of this study was to assess ICD insertion site accuracy and the factors affecting accuracy by emergency centre doctors. Methods This was an observational simulation-based study. Using ultrasound and an invisible ultraviolet marker, the chest wall structures of two simulated patients of different BMI categories (low and high) were marked. Participants placed stickers bilaterally on both patients at their intended ICD insertion site. Sticker placements were photographed under ultraviolet light and analysed. Accurate placement (within the area above the 6th rib, posterior to pectoralis major and 1cm anterior to the mid- axillary line) and within the British Thoracic Society’s ‘triangle of safety’ was noted. Comparisons of accuracy were made between participant rank (junior vs senior), participant experience (<30 vs >30 ICDs placed), BMI category (low vs high) and placement side (left vs right). Results Insertion site was accurate in 47% of placements and within the ‘triangle of safety’ in 51% of placements. Accuracy was associated with greater participant experience (61% vs. 37%, p<0.01), and left-sided placement (54% vs 40%, p = 0.02). No difference was found when comparing patient BMI category (low vs high, 43% vs 51%, p=0.18) or participant rank (junior vs senior, 41% vs 51%, p=0.10). Of all placements, 13% were over muscle or the diaphragm. Conclusion Overall accuracy of ICD insertion site was low, and accuracy was only associated with greater participant experience and left sided placement. Further studies to determine methods to improve ICD placement accuracy are needed. 10 Keywords Chest tubes Thoracostomy Clinical competence Intercostal drain Complications African relevance § High burden of respiratory disease and trauma in low- and middle-income countries results in a high rate of intercostal drain requirement. § Placement of intercostal drains is a critical skill in acute care medicine that is often carried out unsupervised by junior or inexperienced doctors working in overburdened public healthcare systems § Complications associated with misplaced intercostal drains are an unnecessary added strain on already overloaded healthcare systems. § Identification of factors associated with intercostal drain misplacement is an important step in the path to improving accuracy of placement and thus reducing the complications associated with misplacement. 11 Background Intercostal drain (ICD) insertion is a common emergency centre (EC) procedure [1,2]. High rates of penetrating thoracic trauma and HIV-associated pulmonary illness have led to a significant number of patients presenting to the EC requiring ICD insertion in South Africa (SA) [3,4]. Training to insert an ICD correctly is an essential skill for doctors working in acute care [2,5]. SA is a middle-income country which means that in certain hospitals there are staff and resource limitations along with a high patient load. This medical landscape means that junior doctors are expected to perform procedures, such as ICD insertion, which would usually be completed by specialists in more developed countries [2,6–8]. An SA study identified that there was inadequate supervision of junior staff inserting ICDs, wide variability in the experience level of staff regarding ICD insertion and higher complication rates associated with ICD insertion by staff outside of major trauma centres [2]. As with most medical procedures, complications can occur with the insertion of an ICD. The classification thereof is not standardised with complication rates varying between 1-40% [9]. Complications may result in longer hospital stays, additional investigations and reparative surgery for visceral or vascular injury [10]. ICD complication-related morbidity and mortality including financial implications can thus be significant, particularly in a resource-constrained environment [2,9,10]. Several commonly used guidelines describe landmark-based palpation techniques for ICD placement including the British Thoracic Society (BTS), the Advanced Trauma Life Support (ATLS) and the European Trauma Course (ETC) methods [11– 12 13]. BTS guidelines refer to the ‘triangle of safety’ as the appropriate area for ICD insertion. This triangle constitutes: the lateral border of the pectoralis major anteriorly, the fifth intercostal space inferiorly and the lateral border of latissimus dorsi posteriorly, with the axilla as the apex [11]. ATLS guidelines recommend placement in the fourth/fifth intercostal space between the anterior and mid-axillary lines [12]. The ETC guidelines recommend placement one hand width below the anterior axillary fold just anterior to the mid-axillary line [13]. ATLS principles have been adopted and taught as part of a formal ATLS training course in many countries including South Africa, however not all doctors who are expected to place ICDs are certified [7]. The Trauma Society of South Africa recommends the method taught by ATLS, however, the BTS triangle of safety has been described in previous South African studies as convention and used when analysing ICD placement in previously carried out South African studies. [7,8]. It has been shown that it is difficult to find the correct area by palpation alone [13,14]. The aim of this study was to describe the accuracy of palpation-based ICD insertion site identification by EC staff of varying levels of experience and describe some of the factors associated with ICD misplacement. Methodology A prospective observational simulation-based study design was used. The study was approved by the University of the Witwatersrand Human Research Ethics Committee (M1911137). Written informed consent was obtained from all participants and simulated patients. 13 The study population comprised of EC doctors working in three academic hospitals in Johannesburg, SA. Doctors were ranked as either junior or senior. Junior doctors included second year interns and community service medical officers. Senior doctors included medical officers and emergency medicine registrars. Data collection Data were collected from August – September 2020. Two male volunteers acted as simulated patients, one with a body mass index (BMI) of 20 and one with a BMI of 34. The simulated patients were positioned supine with arms behind the head and the head of the bed raised at 30o. Prior to the start of the data collection, ultrasound was used to identify thoracic structures on the simulated patients, which were then marked using an invisible 8280 Special Securitas UV Marker (Edding, Germany). This included the diaphragm at the end of inspiration and expiration, the 4th-7th intercostal spaces, the lateral border of the pectoralis major muscle and the lateral border of the latissimus dorsi muscle (Figure 1A). The chest wall was then divided into four quadrants (A-D) by a line drawn 1cm anterior to the mid-axillary line in a cranio-caudal direction, intersected by a second line along the inferior border of the fifth intercostal space (Figure 1B). Quadrants A and B combined represent the BTS triangle of safety. For this study, Quadrant A represents the ‘accurate’ placement area, anterior to the line 1cm anterior to the mid-axillary line and above or within the 5th intercostal space. Quadrants B-D represent areas of placement that would be deemed not accurate as Quadrants B and C would be too posterior and Quadrants C and D would be too caudal on the thoracic wall. Cadaveric studies that assessed markers placed using the three commonly used palpation techniques (BTS, ATLS, ETC) concluded that not all areas within the BTS ‘triangle of safety’ are, in fact, safe. 14 Sites posterior to the line 1cm anterior to the mid-axillary line place the long thoracic nerve at risk [13]. Although Quadrant B lies within the triangle of safety it is posterior to the line 1cm anterior to the mid-axillary line and thus was not considered to be within the accurate area for this study. Thus, accurate placement in this study only included placements anterior to the line 1cm in front of the mid-axillary line. Figure 1 – Photograph (A) and diagram (B) indicate how the right side of the thoracic wall was demarcated in the simulated patients. Photograph A additionally shows the circular sticker that was used by participants to indicate their preferred ICD insertion site. Diagram B has a green shaded area that represents the expiratory and inspiratory extents of the diaphragm. The anterior light blue line represents the lateral margin of pectoralis major and the posterior light blue line represents the lateral margin of latissimus dorsi. The dashed purple line represents the line 1cm anterior to mid-axillary line. The solid purple line represents the inferior margin of the 5th intercostal space. The red lines that intersect the solid purple line indicate adjacent intercostal spaces’ inferior margin. Quadrants A-D shown as marked in diagram B. Participants used their preferred palpation technique to locate and mark the ICD insertion site on each side of both of the simulated patients’ chest walls with a 15 circular sticker. Participants were unable to see the markings made with the ultraviolet marker and were assessed individually. Each sticker was photographed in situ using a Canon 850D SLR camera (Canon, SA), with the ultraviolet markings made visible using a Zartek ZA-490 UV Flashlight (Zartek, SA). During analysis, marked structures were divided into two groups: muscular (latissimus dorsi and pectoralis major) and diaphragmatic structures (diaphragm/ subdiaphragmatic). Intercostal space placement was divided into 5th intercostal space or above and 6th Intercostal space or below. Statistical Analysis Descriptive variables were presented as frequencies and percentages. The data was analysed using IBM SPSS Statistics, version 22 (2013). The Chi-square test was used to compare frequency of ICD placement between the following variables: rank, experience level, BMI, and side of placement. Significance testing was set at the 95% confidence level with a p-value < 0.05 indicating statistical significance. Results In total, 71 doctors participated, each placing 4 simulated ICDs, resulting in a total of 284 placements. The participant characteristics are shown in Table 1. Table 1: Participant demographics Ranking of doctor Number of ICDs inserted n (%) Junior doctors 27 (38) < 30 ICDs 26 (96) > 30 ICDs 1 (4) 16 Senior doctors 44 (62) < 30 ICDs 15 (34) > 30 ICDs 29 (66) ICD = intercostal drain Accuracy of placement and placement within the triangle of safety Of the 284 simulated ICDs, 47% were placed in the predefined ‘accurate’ area and 51% fell within the BTS triangle of safety. The differences in placement between rank, experience level, patient BMI and side of placement are shown in Table 2 for both the defined accurate area and the BTS triangle of safety. Table 2: Accuracy of placement and placement in BTS triangle of safety Accurate Placement n (%) n (%) p-value Rank Senior doctors Junior doctors 90 (51)# 44 (41) 0.10 Experience level >30 ICDs placed <30 ICDs placed 73 (61) 61 (37) < 0.01 Patient BMI Low BMI High BMI 61 (43) 73 (51) 0.18 Side of placement Left Right 77 (54) 57 (40) 0.02 Placement in BTS Triangle of Safety n (%) n (%) p-value Rank Senior doctors Junior doctors 96 (55) 49 (45) 0.10 Experience level >30 ICDs placed <30 ICDs placed 77 (62) 68 (41) <0.01 Patient BMI Low BMI High BMI 64 (45) 81 (57) 0.04 Side of placement Left Right 17 84 (59) 61 (43) 0.01 BTS = British Thoracic society; ICD = Intercostal drain; BMI = Body mass index Note: percentages reflect proportion of accurate and BTS placements within the indicated group – e.g. 51% of all placements by senior doctors were accurate# Anatomical location of placements based on marked structures Overall, 13% of placements were overlying marked structures: 4% overlying the diaphragmatic structures and 9% overlying the muscular structures. Senior doctors were significantly less likely to place the simulated ICDs over marked structures when compared to junior doctors (10% vs 19%, p = 0.03). Both senior doctors and junior doctors placed 4% of attempts over diaphragmatic structures. Senior doctors placed 6% over muscular structures, compared to 15% of junior doctors (p < 0.01). No significant differences were found regarding placements overlying marked structures when comparing experience level, patient BMI and side of placement. Participants placed 163 attempts (57%) within the 5th intercostal space or above. Almost a quarter of attempts (23%) were closest to the 7th intercostal space. Participants with a higher experience level indicated a significantly larger proportion of simulated ICDs in the 5th intercostal space or above (69% vs 49%, p < 0.01) whereas no significant difference was found when comparing rank, patient BMI or side of placement. Discussion Accurate placement of ICDs in patients is paramount in ensuring appropriate medical management and patient safety. It is thus vital that practitioners inserting ICDs are aware of the importance of accurate placement, as well as the various factors that could possibly contribute to poor placement. 18 The overall accuracy of placement of simulated ICDs in the BTS triangle of safety was low. These low accuracy rates are comparable to those found in related literature. A United Kingdom-based study assessing 50 junior doctors found that only 44% accurately located the triangle of safety, while another clinical audit showed 55% of doctors chose ICD placement inside the BTS triangle of safety [15,16]. ICD placement in the EC is common and placement outside of the accepted ‘safe’ areas is associated with higher complication rates [1,8,11,12]. A South African study by Sritharen et al. noted an overall complication rate of 16% and a meta-analysis by Hernandez et al. found a complication rate of 19% for the insertion of ICDs in trauma patients [9,17]. These significant complication rates as well as those described in other studies emphasise the importance of accurate, safe placement of ICDs [2,7,9,18]. This highlights an urgent need for intervention to improve accuracy of placement and potentially reduce complication rates. However, the rate of accuracy is not akin to the complication rate, as complications associated with depth, ectopic insertion, infection, and angle of insertion can still occur with accurate placement [5,13,17,19]. An institutional audit of placed ICDs would be required to assess complication rates, which could then be compared to accuracy of placement. In contrast to other studies that found a significantly higher complication rate associated with ICDs placed by junior staff, this study found no difference when comparing participant rank [7–9,15]. This discrepancy may be a result of variations in training and clinical experience. A study that assessed anatomical placement of ICDs by Kong et al. found that only 28% of reviewed ICD placements by junior 19 doctors (interns) were located within the BTS triangle of safety [8]. The authors found a significant difference when comparing first and second year intern doctors as well as those who had previously attended an ATLS course [8]. These findings likely reflect improvement in accuracy with experience, which may be reflected in the results from the current study. Self-reported level of experience (number of ICDs placed) may serve as a better proxy than rank, as shown in Table 2. An audit from the United Kingdom also showed previous experience in ICD placement to be a critical factor in choosing the correct placement site [15]. The best approach to improving accuracy would thus likely be a focus on methods that build upon experience, such as simulation training, improved guidance and oversight of placement during internship years and trauma course (e.g ATLS) attendance [8,20]. Anatomical factors such as large body habitus have been reported to be associated with more difficult ICD placement and higher complication rates [1,19]. However, Sethuraman et al. found no significant increase in complications in patients with a large body habitus who had ICDs placed [20]. The results of the current study were not in keeping with the available literature as no statistical difference was found when comparing accurate placement according to patient BMI, although a significant difference was found when assessing placement within the BTS triangle of safety. The accuracy was, however, better in the simulated patient with the higher BMI. This outcome may have been influenced by a possible awareness by the participants of the comparison between low and high BMI as both simulated patients were in the same room. This Hawthorne effect may have resulted in participants taking greater care when placing the sticker in the simulated patient with the higher BMI [21]. 20 The finding of a difference in accuracy and placement in the triangle of safety depending on the side of placement in the patient is a finding not in keeping with other literature either. Carter et al. found no difference when comparing left versus right sides [22]. The significance of this is uncertain. The handedness of participants and the fact that sticker placement is not as complex a motor task as actual ICD placement are possible factors which were not evaluated. This difference would be better assessed by auditing actual ICD insertions. Placements overlying diaphragmatic structures would have significant potential for patient harm and therefore major cost implications in an already resource- constrained setting [2,10]. Avoiding such placement should thus be a particular focus during ICD insertion training. In this study, although it was found that junior doctors were significantly more likely to place the stickers over marked structures, it was more common over muscular and not diaphragmatic structures. That said, placement through large muscles should also be avoided and awareness of these anatomical structures when placing ICDs should be emphasised. The incorporation of ultrasound as an adjunct when planning ICD insertion by junior doctors may ameliorate this [14]. Diaphragmatic excursion has been noted in other studies to be as low as the 7th intercostal space where the diaphragm contacts the costal portion of the parietal pleura in expiration [22]. Thus, peritoneal placement becomes more likely in or below the 7th intercostal space. Kwiatt et el. describe diaphragmatic excursion as rising to as high as the 4th intercostal space in full expiration, thus placement lower than the 5th intercostal space would be a higher risk of subdiaphragmatic placement. This is 21 especially true in pregnant women with a gravid uterus, obese patients and patients with intraabdominal tumours or ascites [1]. In one cadaveric study it was found that more than 80% of the ICDs were placed in the 6th intercostal space or below [23]. Carter et al. assessed marker placement by EC registrars and consultants and found that 36.2% of placements were in the 4th/5th intercostal space [22]. Both subgroups would be classed as senior doctors as per the current study’s criteria. A much higher rate of senior doctors placed the marker in the 5th intercostal space or above. Doctors with more ICD insertion experience were also more likely to place the ICD in the 5th intercostal space or above. This discrepancy between findings may be a function of experience as many ICDs are placed in SA related to the high trauma and tuberculosis caseload [2,22]. Doctors should be aware of factors that may hinder accurate ICD insertion such as female biological sex, high BMI and previous trauma or distorted anatomy [1]. These are the cases where ultrasound-guided ICD placement could be used, or alternatively, senior supervision could be requested [1,14]. A study comparing accuracy of placement within the 5th intercostal space using traditional palpation techniques versus using ultrasound found that only 48% of the attempts by EM residents and students were in the correct intercostal space when using palpation. After a short hands-on training session in the use of ultrasound to identify the intercostal space, the accuracy of placement improved to 91% [14]. Ultrasound- guided placement in non-emergent cases, especially where factors have been identified that may make the insertion of an ICD more difficult, should be considered. 22 Departments that are responsible for many drain placements may also consider a checklist [1]. Checklists could assist junior/inexperienced doctors to prepare correctly for the procedure and potentially make successful placement more likely. A checklist could also help make them aware of potential difficulties and offer solutions such as adequate analgesia and safe sedation methods for the combative patient. Another potential intervention could be the implementation of simulation-based training. A study by Leger et al. showed that simulation-based training improved ICD insertion success rate in a traumatic pneumothorax model [24]. This could be explored as a method to improve upon experience without having to place ICDs in a clinical setting. Study limitations This was a small study with 71 participants: a small proportion of the number of doctors working within ECs in Johannesburg. Of these, few were junior doctors as the majority of EC staffing is made up of medical officers. This was a simulation- based situation and would thus not reflect the pressures involved during actual ICD insertion in an EC. The simulation setting may have resulted in participants taking greater care when placing the simulated ICD in the simulated patient with the higher BMI as part of a Hawthorne effect [21]. This study included three large academic hospitals in Johannesburg, as such there was no assessment of smaller facilities, rural facilities, or facilities in other parts of SA. Two of the centres assessed were medical emergency units which are separate departments from the trauma units. The non-trauma ECs receive fewer trauma patients, which constitutes a large proportion of patients who require ICD insertion in South Africa [2]. Doctors in those departments were less likely to be as experienced in inserting ICDs than those doctors from the trauma units. Due to anonymity, this could not be analysed 23 separately. The simulated patients were requested to remain with the top half unclothed for the duration of each session, no female simulated patients were thus incorporated in the study. Assessing female breast tissue as a possible contributing factor to ICD misplacement was thus not carried out as a part of this study. Conclusion Accurate placement of ICDs is imperative to reduce the rate of complications associated with ICD insertion. The overall accuracy of placement was worryingly low in this simulation-based study. The clinical experience of the inserting doctor was found to be a significant factor affecting the accuracy of ICD placement. Experience was found to be a common theme throughout this study and the associated literature. Methods to improve upon experience and thus accuracy should be further studied, and potentially include simulation-based training and ultrasound-guided placement. Training should emphasise the importance of recognising factors predicting difficult ICD insertion. Senior staff should be encouraged to oversee junior staff ICD placement so that immediate corrective action can be undertaken to prevent patient harm. 24 References [1] Kwiatt M, Tarbox A, Seamon MJ, Swaroop M, Cipolla J, Allen C, et al. Thoracostomy tubes: a comprehensive review of complications and related topics. Int J Crit Illn Inj Sci 2014;4:143–55. https://doi.org/10.4103/2229-5151.134182. [2] Maritz D, Wallis L, Hardcastle T. Complications of tube thoracostomy for chest trauma. S Afr Med J 2009;99:114–7. [3] Norman R, Matzopoulos R, Groenewald P, Bradshawa D. The high burden of injuries in South Africa. Bull World Health Organ 2007;85:695–702. https://doi.org/10.2471/BLT.06.037184. [4] Das S, Ghoshal B. Pneumothorax in human immunodeficiency virus infection. JACP 2015;3:38–40. https://doi.org/10.4103/2320-8775.158834. [5] Jones CW, Rodriguez RD, Griffin RL, McGwin G, Jansen JO, Kerby JD, et al. Complications associated with placement of chest tubes: a trauma system perspective. J Surg Res 2019;239:98–102. https://doi.org/10.1016/j.jss.2019.01.012. [6] The World Bank. South Africa – World Bank data, https://data.worldbank.org/country/ZA; 2022 [Accessed 10 February 2022]. [7] Kong VY, Oosthuizen G, Sartorius B, Keene C, Clarke DL. An audit of the complications of intercostal chest drain insertion in a high volume trauma service in South Africa. Ann R Coll Surg Engl 2014;96:609–13. https://doi.org/10.1308/003588414X14055925058599. [8] Kong VY, Oosthuizen G, Sartorius B, Keene CM, Clarke DL. Correlation between ATLS training and junior doctors’ anatomical knowledge of intercostal chest drain insertion. J Surg Educ 2015;72:600–5. https://doi.org/10.1016/j.jsurg.2015.01.022. [9] Sritharen Y, Hernandez MC, Haddad NN, Kong V, Clarke D, Zielinski MD, et al. External validation of a tube thoracostomy complication classification system. World J Surg 2018;42:736–41. https://doi.org/10.1007/s00268-017-4260-8. [10] Hernandez MC, Zeb MH, Heller SF, Zielinski MD, Aho JM. Tube thoracostomy complications increase cost. World J Surg 2017;41:1482–7. https://doi.org/10.1007/s00268-017-3897-7. [11] Havelock T, Teoh R, Laws D, Gleeson F. Pleural procedures and thoracic ultrasound: British Thoracic Society pleural disease guideline 2010. Thorax 2010;65:i61–76. https://doi.org/10.1136/thx.2010.137026. 25 [12] American College of Surgeons Committee on Trauma. Advanced Trauma Life Support - Student Course Manual. 10th Ed. Chicago, IL: American College of Surgeons; 2018. [13] Bowness J, Kilgour PM, Whiten S, Parkin I, Mooney J, Driscoll P. Guidelines for chest drain insertion may not prevent damage to abdominal viscera. Emerg Med J 2015;32:620–5. https://doi.org/10.1136/emermed-2014-203689. [14] Taylor LA, Vitto MJ, Joyce M, Tozer J, Evans DP. Ultrasound-guided thoracostomy site identification in healthy volunteers. Crit Ultrasound J 2018;10:1–5. https://doi.org/10.1186/s13089-018-0108-1. [15] Griffiths JR, Roberts N. Do junior doctors know where to insert chest drains safely? Postgrad Med J 2005;81:456–8. https://doi.org/10.1136/pgmj.2004.024752. [16] Elsayed H, Roberts R, Emadi M, Whittle I, Shackcloth M. Chest drain insertion is not a harmless procedure - are we doing it safely? Interact Cardiovasc Thorac Surg 2010;11:745–8. https://doi.org/10.1510/icvts.2010.243196. [17] Hernandez MC, El-Khatib M, Prokop L, Zielinski MD, Aho JM. Complications in tube thoracostomy: systematic review and meta-analysis. J Trauma Acute Care Surg 2018;85:410–6. https://doi.org/10.1097/TA.0000000000001840. [18] Kong VY, Clarke DL. The spectrum of visceral injuries secondary to misplaced intercostal chest drains: Experience from a high volume trauma service in South Africa. Injury 2014;45:1435–9. https://doi.org/10.1016/j.injury.2014.05.013. [19] Hernandez MC, Laan D, Zimmerman SL, Naik ND, Schiller HJ, Aho JM. Tube thoracostomy: increased angle of insertion is associated with complications. J Trauma Acute Care Surg 2016;81:366–70. https://doi.org/10.1097/TA.0000000000001098. [20] Sethuraman KN, Duong D, Mehta S, Director T, Crawford D, George JS, et al. Complications of tube thoracostomy placement in the emergency department. J Emerg Med 2011;40:14–20. https://doi.org/10.1016/j.jemermed.2008.06.033. [21] Sedgwick P, Greenwood N. Understanding the Hawthorne effect. BMJ 2015;351:h4672–3. https://doi.org/10.1136/bmj.h4672. [22] Carter P, Conroy S, Blakeney J, Sood B. Identifying the site for intercostal catheter insertion in the emergency department: is clinical examination reliable? Emerg Med Australas 2014;26:450–4. https://doi.org/10.1111/1742-6723.12276. 26 [23] Bowness JS, Nicholls K, Kilgour PM, Ferris J, Whiten S, Parkin I, et al. Finding the fifth intercostal space for chest drain insertion: guidelines and ultrasound. Emerg Med J 2015;32:951–954. https://doi.org/10.1136/emermed-2015-205222. [24] Léger A, Ghazali A, Petitpas F, Guéchi Y, Boureau-Voultoury A, Oriot D. Impact of simulation-based training in surgical chest tube insertion on a model of traumatic pneumothorax. Adv Simul (Lond) 2016;1:1–7. https://doi.org/10.1186/s41077-016- 0021-2. 27 Section 2 – Research Protocol A DESCRIPTIVE STUDY OF THE ACCURACY OF INTERCOSTAL DRAIN PLACEMENT AND FACTORS INFLUENCING PLACEMENT. Dr Tessa Korda MMed Emergency Medicine Student Number: 433192 Staff Number: A0032101 Supervisors: Dr T. Baillie-Stanton, Prof. L.N. Goldstein INTRODUCTION Background A large number of intercostal drains (ICDs) are placed in patients in South Africa (1). The high levels of interpersonal violence and motor vehicle collisions as well as the scourge of Human Immunodeficiency Virus and its associated respiratory complications have led to significant pathology presenting to the emergency department (ED) requiring ICD insertion(2, 3). South Africa is a middle-income country where the public health system has many resource restrictions(4). The resource limitations within the healthcare system extend to staff numbers with many EDs being sub-optimally staffed(2). Staff and physical resource limitations along with high patient numbers have created a medical landscape whereby junior medical doctors are expected to perform a number of procedures that would only be completed by seniors or even specialists in many developed countries(5). Rural areas are more likely to be staffed by junior doctors, who would likely be expected to perform invasive procedures such as ICD insertion with minimal supervision. It has been reported that there is inadequate supervision and guidance of junior staff inserting ICDs in South Africa. A study by Maritz et al.(1) found that there were higher 28 complication rates associated with ICDs placed by staff outside of major trauma centres. There is wide variation in the level of expertise and experience of staff responsible for ICD insertion in South Africa(1). This variation may be a factor when considering complications associated with ICD insertion. Complications of ICD insertion Complications can be classified as insertional, positional, related to removal, infective/immunologic as well as instructional/educational/equipment-related(6). The insertion of intercostal drains in the incorrect anatomical space is associated with a range of complications, including diaphragmatic, visceral and vascular injury. Complications associated with malposition are the most common and incorrect placement can result in significant morbidity and mortality(1, 7). The rates of complications associated with ICD insertion are significant(6). A UK-based study showed 45% of junior doctors placed the ICD outside of the accepted safe zone(8). In a South African study Sritharen et al., they noted that 44% of documented complications were positional. In that study, complications occurred in 16% of the patients’ ICD placements analysed(6). Implications of complications in a resource-restricted setting such as SA are magnified due to an already financially-strained healthcare system. The cost to the healthcare system as a result of ICD-associated complications can be high including further resource utilisation associated with correction of a misinserted drain(1). A CT scan may be required in any patient where drain malposition is suspected(5). Patients with complications would require a longer hospital stay and reparative surgery may be required in cases of visceral or vascular injury. 29 Methods for determining ICD insertion site There are multiple methods of determining ICD insertion site. The method used may affect the accuracy of placement. There are two well-recognised techniques: a landmark-based palpation technique and an ultrasound-assisted technique. Landmark-based palpation techniques of placement are commonly used. There are several guidelines that have been published and used in this regard: The Advanced Trauma Life Support (ATLS) method, the European Trauma Course method and the British Thoracic Society guidelines method. According to these guidelines, the ideal area for ICD insertion in order to avoid ICD associated complications is the 4th intercostal space 1cm anterior to the mid-axillary line(9, 10). In many patients it has been shown to be difficult to find the correct point by palpation alone(11). Applying these guidelines in cadaveric studies found that more than 80% of the ICDs were being placed in the 6th intercostal space or below(10). British Thoracic society guidelines refer to the ‘triangle of safety’ as the anatomical area in which an ICD should be placed. This triangle is defined as: an anatomical triangle bounded by the lateral border of the pectoralis major anteriorly, the fifth intercostal space inferiorly and the lateral border of latissimus dorsi laterally with the axilla as the apex(5). The ultrasound-assisted placement technique has been used in a number of settings where non-emergent ICD placement occurs. Ultrasound-assistance has been shown to reduce the complication rate of pleural procedures in the critical care setting thus its use has been recommended in the British Thoracic Society pleural disease guidelines(5). Ultrasound has also been shown to aid in prompt identification of a misinserted ICD(7). 30 Factors contributing to ICD misinsertion A number of factors have been found to contribute to misinsertion of an ICD. These factors include patient factors and staff factors. Some of the patient factors include patient habitus (increased body mass index (BMI)); female biological sex; previous chest drain insertion; rib trauma; intra-abdominal pathology; and parenchymal respiratory disease(12). A larger body habitus makes identification of the anatomical site for insertion more difficult and is noted to be associated with multiple types of complications related to insertion and position(12). The use of the ATLS method of placement was shown in one study to cause a higher rate of insertional complications in women(10). In patients who have had previous chest procedures or ICDs as well as those with significant anatomical distortion secondary to trauma or respiratory disease, ICD drain insertion may be more difficult. Staff factors include the level of experience of the particular individual, the quality of training received for the insertion of ICDs and prior ATLS accreditation(1, 13). Summary Misinsertion of ICDs occurs frequently in South Africa. This has significant patient as well as cost implications. There are a number of factors associated with misinsertion. The aim of this study is to describe the accuracy of palpation-based ICD insertion site identification by ED staff of varying levels of experience and to assess some of the factors associated with ICD misinsertion. Study Objectives • To describe the overall accuracy of landmark-based palpation techniques for site identification for ICD placement by ED staff of varying levels of 31 experience on the University of the Witwatersrand Emergency Medicine circuit. • To compare the accuracy of site choice between staff of different levels of experience in two major groups: Junior doctors (second year interns and community service doctors) versus Senior doctors (medical officers and registrars) • To compare the accuracy of site choice between a simulated patient with a normal BMI of 20-24 and a simulated patient with a BMI of more than 30 • To describe any other factors staff identify as potentially contributing to increased difficulty in site identification. Study design This is a prospective, cross-sectional observational study describing the accuracy of site choice and factors that influence the accurate siting of insertion of an ICD by ED staff from three university-affiliated academic hospitals in Johannesburg. Study population and Sample The study population will include registrars from the University of Witwatersrand Division of Emergency Medicine as well as the second-year medical interns, community service officers and medical officers working in three academic EDs in Johannesburg: Helen Joseph Hospital, Chris Hani Baragwanath Academic Hospital and Charlotte Maxeke Johannesburg Academic Hospital. Inclusions: All doctors in the abovementioned EDs who consent to participate at the pre-arranged data collection sessions on an ad-hoc basis. 32 Sample size: A total of 60 participants will be assessed. There will be 30 participants in the junior category (second year interns and community service officers) and 30 participants in the senior category (medical officers and registrars) Detailed Description of Methods and Techniques Prearranged sessions will be held in each of the three EDs, as well as sessions attached to registrar teaching days. Two simulated patients will be used for each session – one with a normal habitus (BMI 20-24) and one with an obese habitus (BMI >30). The simulated patients will be positioned in a standard manner for ICD insertion: supine in a semi-recumbent position between 30 and 45 degrees, arms abducted and hands placed behind the head. The bed and patient position on the bed will be meticulously noted during ultrasound identification of anatomical structures so that the overlying skin markings are maintained in the correct position during participant recordings. Prior to each session the fourth intercostal space, the triangle of safety and pertinent thoracic structures on the left and right will be located using an ultrasound machine. The available ultrasound in each ED will be used to delineate the appropriate landmarks as follows: The linear probe with the marker pointing in a cranial direction will be used to identify the clavicle and then the second rib, which correlates to the first rib shadow below the clavicle. Thereafter the second rib space between the second and third rib shadows will be identified, the probe will then be moved in a caudal direction, perpendicular to the skin, until the fourth intercostal space is identified. The fourth intercostal space will then be followed to the mid-axillary line. The UV marker will be used to mark the 4th and 5th intercostal spaces, smaller 33 secondary marks will be drawn in the other intercostal spaces along the mid-axillary line so they may be identified more easily. The triangle of safety will be delineated with the UV marker (lateral border of pectoralis major, anterior border of latissimus dorsi, 5th intercostal space) and marked with an “8280 Special Securitas UV Marker” (Edding, Germany) that will be invisible to the participants. The diaphragmatic border will be marked in both inspiration and expiration. This will be performed on each simulated patient. Once all of the relevant thoracic structures have been marked the lateral thoracic wall will be divided into four quadrants: a line will be drawn 1cm anterior to the midaxillary line in cranial to caudal direction. This line will be intersected by a line drawn along the inferior border of the fifth intercostal space to form the four quadrants labelled A, B, C and D as seen in figure 1. Figure 2 Quadrants as Drawn on the lateral chest wall Once consent has been signed, the participants will have an 4-minute period during which they will be asked to place four stickers (2 per simulated patient). One sticker will be placed on each side of each patient to indicate the participant’s chosen ICD insertion site according to their usual palpation-based method for site identification. After marking the simulated patients, each participant will also be given a short questionnaire (Appendix A) to complete. 34 After each participant has placed their four stickers, photographs will be taken using a “Canon 800d” (Canon, Japan) single lens reflex camera and a tripod. The camera will be set up prior to each session ensuring the lens (18-55mm) is at the same height as and perpendicular to the simulated patients mid-lateral thorax. A “ZA-490 UV Flashlight, 9 LED” (Zartek, South Africa) black light will be used to illuminate the pre-marked ideal point, this will be photographed along with the participant sticker, a standardised scale marker, and an information marker indicating the unique study number, group designation (junior/senior), side (left/right) and body habitus category for each photograph. A participant sticker will be regarded as accurately placed if it is in the fourth or fifth intercostal space anterior to the mid-axillary line and posterior to the lateral border of the pectoralis major muscle within ‘quadrant A’. The intercostal space closest to the participant sticker will be recorded on the data collection sheet (Appendix B). The quadrant within which the participant sticker is placed will also be recorded. Variables The quadrant within which the participant sticker falls. The closest intercostal space to each sticker will be recorded. Variables collected using the data sheet will include the participant’s current rank and approximate number of ICD’s previously inserted. 35 Statistical Analysis The overall accuracy according to distance from correct ICD placement site for all participants will be described using medians and inter-quartile ranges. ICD placement will be categorised into correct and incorrect placement. The following will then be compared in order to check for differences:- • doctor experience (junior/senior group) • patient body habitus category (normal/obese) • side (left/right) The answers to the open question at the end of the questionnaire will be categorised and described. The actual placements will be superimposed and presented as a figure for each simulated patient. Limitations Ideally a larger array of simulated patients with different factors influencing difficulty of ICD insertion should be studied e.g. Female with large breasts, tall vs short, geriatric patients etc. Ethics and Permissions An application will be made to the University of the Witwatersrand Human Research Ethics Committee (Medical) for approval. Permission will be requested from the relevant authorities and heads of department at the aforementioned hospitals. 36 Timing Funding The study will be self-funded. A predicted budget including a UV light and pen as well as subsidizing transport for simulation patients is approximately R2500. UV Light: R300 Batteries: R80 UV markers: R171 (3) Paper and Printing: R180 Simulated patient subsidy: 6 sessions x R150 x 2 = 1800 FEB MAR APR MAY JUNE JULY AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUNE JULY AUG SEP LIT. R/V PROTOCOL PREP PROTOCOL ASSESSMENT ETHICS APPROVAL DATA COLLECTION DATA ANALYSIS WRITE UP THESIS 37 References 1. Maritz D, Wallis L, Hardcastle T. Complications of tube thoracostomy for chest trauma. South African Medical Journal. 2009;99(2):114-7. 2. Norman R, Matzopoulos R, Groenewald P, Bradshaw D. The high burden of injuries in South Africa. Bulletin of the World Health Organization. 2007;85(9):695-702. 3. Das SK, Ghoshal B. Pneumothorax in human immunodeficiency virus infection. The Journal of Association of Chest Physicians. 2015;3(2):38-40. 4. The World Bank. Data for South Africa, Middle income 2017 [Date accessed: 2019- 09-10] Available from: https://data.worldbank.org/?locations=ZA-XP. 5. Havelock T, Teoh R, Laws D, Gleeson F. Pleural procedures and thoracic ultrasound: British Thoracic Society pleural disease guideline 2010. Thorax. 2010;65(Suppl 2):i61-i76. 6. Sritharen Y, Hernandez M, Haddad N, Kong V, Clarke D, Zielinski M, et al. External Validation of a Tube Thoracostomy Complication Classification System. World Journal of Surgery. 2018;42(3):736-41. 7. Menegozzo CAM, Utiyama EM. Steering the wheel towards the standard of care: Proposal of a step-by-step ultrasound-guided emergency chest tube drainage and literature review. International Journal of Surgery. 2018;56:315-9. 8. Griffiths JR, Roberts N. Do junior doctors know where to insert chest drains safely? Postgraduate Medical Journal. 2005;81(957):456-8. 9. Bowness J, Kilgour PM, Whiten S, Parkin I, Mooney J, Driscoll P. Guidelines for chest drain insertion may not prevent damage to abdominal viscera. Emergency Medicine Journal. 2015;32(8):620-5. 38 10. Bowness JS, Nicholls K, Kilgour PM, Ferris J, Whiten S, Parkin I, et al. Finding the fifth intercostal space for chest drain insertion: guidelines and ultrasound. Emergency Medicine Journal. 2015;32(12):951-4. 11. Taylor LA, Vitto MJ, Joyce M, Tozer J, Evans DP. Ultrasound-guided thoracostomy site identification in healthy volunteers. Critical Ultrasound Journal. 2018;10(1):28-. 12. Kwiatt M, Tarbox A, Seamon MJ, Swaroop M, Cipolla J, Allen C, et al. Thoracostomy tubes: A comprehensive review of complications and related topics. International Journal of Critical Illness & Injury Science. 2014;4(2):143-55. 13. Kong VY, Oosthuizen G, Sartorius B, Keen CM, Clarke DL. Correlation Between ATLS Training and Junior Doctors’ Anatomical Knowledge of Intercostal Chest Drain Insertion. Journal of Surgical Education. 2015;72(4):600-5. 39 APPENDIX A QUESTION SHEET STUDY PARTICIPANT NUMBER_______________________ 1) PLEASE TICK APPROPRIATE BOX: REGISTRAR MEDICAL OFFICER COMM. SERVICE INTERN 2) APPROXIMATE NUMBER OF ICDs PLACED <30 >30 3) HAVE YOU EVER MISPLACED OR WITNESSED AN INTERCOSTAL DRAIN BEING MISPLACED? YES NO IF YES, IS THERE ANY FACTOR THAT YOU RECALL MAKING THE PROCEDURE MORE DIFFICULT THAN USUAL ________________________________________________________ ________________________________________________________ ________________________________________________________ ________________________________________________________ 40 APPENDIX B DATA COLLECTION SHEET STUDY PARTICIPANT NUMBER_______________________ MEASUREMENTS: Left - normal WITHIN TRIANGLE OF SAFETY: Y/N CLOSEST RIB SPACE/RIB SPACE:____________________ ACCURATE PLACEMENT:________________________________ QUADRANT: OVERLYING ANY MARKED STRUCTURES: Right - normal WITHIN TRIANGLE OF SAFETY: Y/N CLOSEST RIB SPACE/RIB SPACE:____________________ ACCURATE PLACEMENT:________________________________ QUADRANT: OVERLYING ANY MARKED STRUCTURES: Left – high BMI WITHIN TRIANGLE OF SAFETY: Y/N CLOSEST RIB SPACE/RIB SPACE:____________________ ACCURATE PLACEMENT:________________________________ QUADRANT: OVERLYING ANY MARKED STRUCTURES: Right – high BMI WITHIN TRIANGLE OF SAFETY: Y/N CLOSEST RIB SPACE/RIB SPACE:____________________ ACCURATE PLACEMENT:________________________________ QUADRANT: OVERLYING ANY MARKED STRUCTURES: 41 Annexures Annexure A – HREC clearance certificate 42 Annexure B – Turn-it-in plagiarism report 43 Annexure C – Certificate of submission signed by student 44 45 Annexure D – Study site permissions 46 47 48 49 50 51 Annexure E – Separate copy of abstract Abstract Background South Africa’s significant burden of trauma and respiratory disease requires a high rate of intercostal drain (ICD) insertions by emergency centre practitioners. ICD misplacement is associated with significant complications. The aim of this study was to assess ICD insertion site accuracy and the factors affecting accuracy by emergency centre doctors. Methods This was an observational simulation-based study. Using ultrasound and an invisible ultraviolet marker, the chest wall structures of two simulated patients of different BMI categories (low and high) were marked. Participants placed stickers bilaterally on both patients at their intended ICD insertion site. Sticker placements were photographed under ultraviolet light and analysed. Accurate placement (within the area above the 6th rib, posterior to pectoralis major and 1cm anterior to the mid- axillary line) and within the British Thoracic Society’s ‘triangle of safety’ was noted. Comparisons of accuracy were made between participant rank (junior vs senior), participant experience (<30 vs >30 ICDs placed), BMI category (low vs high) and placement side (left vs right). Results Insertion site was accurate in 47% of placements and within the ‘triangle of safety’ in 51% of placements. Accuracy was associated with greater participant experience (61% vs. 37%, p<0.01), and left-sided placement (54% vs 40%, p = 0.02). No difference was found when comparing patient BMI category (low vs high, 43% vs 51%, p=0.18) or participant rank (junior vs senior, 41% vs 51%, p=0.10). Of all placements, 13% were over muscle or the diaphragm. Conclusion Overall accuracy of ICD insertion site was low, and accuracy was only associated with greater participant experience and left sided placement. Further studies to determine methods to improve ICD placement accuracy are needed. 52 Annexure F – Supervisor submission form 53 54 Annexure G – ETD Form 55 56 57 58