Abstract
Background
Penetrating pneumothorax and haemothorax are common injuries seen in South African trauma centres. Hospitals in the Western Cape Province of South Africa utilize a protocol for the expedited management of patients with chest injuries requiring intercostal drains (ICD) within emergency department observation units, with emphasis on early mobilization and pulmonary rehabilitation. The study objective was to evaluate the safety and effectiveness of the Western Cape ICD protocol for unilateral traumatic pneumothorax, haemothorax, or pneumo-haemothorax.
Methods
This was a prospective, observational, multi-centre cohort study that enrolled adult (≥ 18 years) patients within four public hospitals (primary, secondary, tertiary) in the Western Cape Province from 2024 to 2025. Patients were eligible for inclusion if they received ICD protocol care at any of the participating hospitals. We performed a descriptive analysis of patient injury profiles, complications, hospital length of stay, in-hospital mortality and 30-day readmissions. The primary outcome was hospital length-of-stay with a hypothesis that > 50% of patients will be discharged ≤ 72 h.
Results
497 patients were enrolled, with 92% male, and median age 29.2. Most (87.7%) were injured by stabbing/cutting mechanisms, and 53.5% had a haemothorax. There were procedural complications among 10.3% of cases (n = 51), including deep placement (n = 25), shallow placement (n = 8), kinked ICD (n = 8), sub-optimal intrapleural placement (too cephalad/caudad; n = 4) and extra-pleural location (n = 4). There were delayed complications among 7.6% of the cohort (n = 38). Common delayed complications included ICD dislodgment (n = 9), premature ICD removal (n = 6), retained haemothorax (n = 6), and infection (n = 15, including 3 empyema, 5 pneumonia, and 9 superficial ICD site infection). Hospital length of stay was a median of 3 days (IQR 2–4). There were no in-hospital deaths, and only 14 patients (2.8%) were readmitted within 30 days.
Conclusion
Our cohort had a composite complication rate of 16.7% at the patient-level, combining malpositioning and delayed complications, suggesting that the Western Cape protocol is safe compared to standard of care for the treatment of simple traumatic pneumothorax and haemothorax.
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1186/s12245-026-01268-0.
Keywords: Africa South of the Sahara, Trauma (wounds and injuries), Emergency service (hospital), Accident and emergency medicine
Introduction
Chest injuries are a common source of morbidity and mortality among trauma victims [1]. Approximately one third of patients with chest trauma have a pneumothorax or haemothorax (PHTX), and many of these patients receive intercostal drains (ICD), also known as chest tubes, for evacuation of blood and gas from the pleural space [2]. The treatment of thoracic trauma continues to evolve, and the current paradigm of ICD care remains laborious and time intensive [3, 4].
Standard management of traumatic PHTX involves ICD insertion with connection to an underwater seal drainage system, which is then typically managed by a pulmonary or surgical service on an inpatient floor. Some health systems manage small spontaneous or traumatic pneumothoraces on an outpatient basis with the use of pleural vents or Heimlich valves [5]. However, there have been cases of treatment failure and adverse events such as tube dislodgment or worsening pneumothorax, which may make clinicians reluctant to discharge patients with such devices in place [6, 7]. An alternative option, in which patients receive ICDs and subsequently remain under observation within the emergency department, promotes patient safety, conservation of inpatient beds, and potential health system cost savings.
South African trauma centres care for large volumes of patients with penetrating PHTX secondary to interpersonal violence [8–10]. Clinicians in these settings have developed innovative techniques to expedite the care of PHTX patients requiring ICD. The Western Cape ICD protocol, developed by clinicians working in the public hospital system surrounding Cape Town, is one such example (Fig. 1). Patients treated under this protocol are adults with simple traumatic unilateral PHTX, who are hemodynamically stable and ambulatory. After surgical (non-pigtail) ICD placement, patients are monitored within the emergency department (ED) observation unit, where they receive oral analgesics (typically a combination of ibuprofen, acetaminophen, and tramadol), vital sign and haemoglobin monitoring. They do not receive wall suction or oxygen therapy unless they fail to progress as expected.
Fig. 1.
Western Cape ICD protocol
Patients’ ICDs are connected to a portable drain system with a one-way dry seal valve; this allows for unrestricted patient movement while the ICD is still in place. The Sinapi Biomedical brand ICD device used in this study was previously studied in a randomized controlled trial at Tygerberg Hospital in South Africa, in which it was compared to traditional underwater drains; it was found to reduce drainage time with similar safety profile [11]. An important aspect of the Western Cape ICD protocol is the emphasis on early mobilization. Patients are given positive end-expiratory pressure devices (PEEP; e.g., a bottle with straw that patients blow bubbles into), and are encouraged to walk, climb stairs, use a stationary exercise bicycle, and/or work with a physical therapist. Patients mobilize to varying degrees depending on staffing resources, patient census, and patient motivation. Patients successfully pass the protocol and undergo ICD removal once they drain less than 50 mL of blood over 24 h and have no evidence of air leak. The goal is safe discharge within three hospital days.
The objective of this study is to assess the safety and effectiveness of the Western Cape ICD protocol using observational data. We describe the injury profile, length-of-stay, and complication rates for patients treated under this protocol. This information may prove useful to clinicians in resource-constrained settings caring for patients with penetrating chest trauma requiring ICD placement.
Methods
Design and approach
This was a prospective, observational, multi-centre cohort study that took place in the Western Cape Province of South Africa over 12 months from March 2024 to February 2025. This study was nested within an ongoing longitudinal observational study titled the Epidemiology and Outcomes of Prolonged Trauma Care (EpiC) [12]. EpiC enrols patients ≥ 18 who present to a study site within 24 h of injury and excludes prisoners and injuries from bites, stings or envenomation.
This study enrolled patients at four public hospitals, including one tertiary academic referral centre, one regional hospital and two district hospitals, all with 24-hour emergency care. Patients were eligible if they met the following criteria: EpiC study eligible, diagnosed with unilateral PHTX, received ICD, able to ambulate, and deemed by treating clinician to be appropriate for routine care under the Western Cape ICD protocol (Fig. 1). Patients were excluded for the following criteria: massive haemothorax, defined as drainage > 1000 mL upon ICD insertion or any autologous transfusion of blood; respiratory failure, defined as intubation or positive pressure ventilation within 24 h of hospital arrival; or high complexity, defined as intensive care unit (ICU) admission, operating theatre intervention for any indication, or death within 24 h of hospital arrival. We excluded patients who were transferred to non-study facilities during their hospitalization.
Study variables were abstracted from hospital, emergency medical services, and forensic pathology records, directly into the EpiC study Research Electronic Data Capture (REDCap) password-protected database. This study received ethics approval from Stellenbosch University Health Research Ethics Committee (Reference N20/03/036) and was conducted in accordance with the Declaration of Helsinki. The Defense Health Agency’s Office of Human Research Oversight, the funder’s ethics committee, provided an additional review and determined compliance with the U.S. Department of Defense’s human subject research protections (OHRO Log Number E01863.1x).
Data analysis
The primary outcome was hospital length-of-stay, with a hypothesis that > 50% of patients will be discharged ≤ 72 h. Secondary outcomes were complication rates, including procedural complications related to ICD placement, along with delayed complications, including ICD dislodgment, retained haemothorax, respiratory failure, infections, and 30-day readmissions. These outcomes were determined through chart review of the medical records. 30-day readmissions were identified using two databases from the Western Cape Department of Health, which capture patient encounters across public healthcare facilities.
Descriptive statistics are presented as frequencies and percentages for categorical outcomes and median [interquartile range, IQR]) for continuous outcomes. The hospital length of stay hypothesis was tested using a one-sample z-test for proportions, with the null hypothesis that proportion discharged ≤ 72 h is ≤ 50%, using a 0.05 level of significance (α). Bivariate logistic regression was used to compare patient groups based on hospital length-of-stay (≤ 72 h versus > 72 h).
Results
There were 497 patients who received ICD care under the Western Cape protocol who met study eligibility criteria (Fig. 2). The cohort was 92% male with median age of 29.2 (IQR 23.9–35.8) (Table 1). Most patients were injured by stabbing/cutting mechanisms (87.7%), with the remainder injured by firearms (6.2%), strikes/hits (4.6%), vehicular injuries (1%), and falls (0.4%). Few patients (n = 57, 11.5%) had trauma outside of the thorax, and overall injury severity was low (median New Injury Severity Score 10, IQR 6–17). Patients enrolled in the protocol had a normal initial systolic blood pressure (median 117, IQR 102–132) and most (n = 397, 79.9%) had an oxygen saturation > 90% on room air upon facility arrival. Few received blood products (n = 26, 5.2%) or tranexamic acid (n = 69, 13.9%).
Fig. 2.
Enrolment diagram. *Patient injury profile too complex for ICD protocol per clinician discretion
Table 1.
Demographics and injury characteristics of patients with simple pneumothorax/haemothorax requiring ICD insertion
| Characteristics | All (n = 497) |
|---|---|
| Male, n (%) | 457 (92.0%) |
| Age, median (IQR) | 29.2 (23.9, 35.8) |
| Comorbidities, n (%) | |
| Diabetes Mellitus | 2 (0.4%) |
| HIV/AIDS | 40 (8.0%) |
| Hypertension | 19 (3.8%) |
| Tuberculosis | 8 (1.6%) |
| Mechanism of Injury, n (%) | |
| Firearm | 31 (6.2%) |
| Struck/hit | 23 (4.6%) |
| Stabbing or cut | 436 (87.7%) |
| Vehicular Injury | 5 (1.0%) |
| Fall | 2 (0.4%) |
| Maximum AIS Thorax Severity Score, median (IQR) | 3.0 (2.0, 3.0) |
| Polytrauma (AIS > 1 non-thorax), n (%) | 57 (11.5%) |
| NISS, median (IQR) | 10.0 (6.0, 17.0) |
| Initial SBP, median (IQR) | 117.0 (102.0, 132.0) |
| Initial SpO2 > 90% on arrival, n (%) | 397 (79.9%) |
| Initial Glasgow Coma Scale, median (IQR) | 15.0 (15.0, 15.0) |
| Any blood product administration within 24 h of injury, n (%) | 26 (5.2%) |
| Any TXA administration within 24 h of injury, n (%) | 69 (13.9%) |
LOS: length of stay; IQR: Inter-Quartile Range; HIV/AIDS: Human Immunodeficiency Virus/Acquired Immunodeficiency Syndrome; AIS: Abbreviated Injury Scale; NISS: New Injury Severity Score; SBP: systolic blood pressure; SpO2: blood oxygen saturation; TXA: tranexamic acid
More patients had a haemothorax (n = 265, 53.5%) compared to pneumothorax (n = 232, 46.7%) (Table 2). A majority (n = 270, 54.3%) were diagnosed by clinical exam alone, with the remainder diagnosed with the assistance of thoracic ultrasound or chest x-ray. Few (n = 11, 2.2%) were diagnosed by computed tomography. Tension physiology was rare (n = 22, 4.4%). The sterility of the ICD insertion was often not documented (unknown for n = 301, 61%), and prophylactic antibiotics were administered for less than one in five cases (n = 95, 19.1%). The initial ICD was malpositioned in 10.3% of cases (n = 51), and most of these (n = 47) were either repositioned or replaced. Reasons for malpositioning were as follows: the ICD was placed too deep (n = 25), too shallow (n = 8), was kinked (n = 8), sub-optimally positioned within the pleural space (n = 4, 2 cases too cephalad, 2 cases too caudad), or was outside of the pleural space (n = 4), including 2 placed intra-abdominally.
Table 2.
Injury characteristics and ICD insertion details
| Insertion Details | All (n = 497) |
|---|---|
| Thoracic injury type, n (%) | |
| Pneumothorax | 232 (46.7%) |
| Haemothorax/Hemopneumothorax | 265 (53.3%) |
| Laterality, n (%) | |
| Left | 270 (54.5%) |
| Right | 225 (45.5%) |
| Diagnostic tools used for PHTX diagnosis, n (%) | |
| Thoracic US | 79 (15.9%) |
| Chest x-ray | 180 (36.2%) |
| CT | 11 (2.2%) |
| Clinical exam only | 270 (54.3%) |
| Tension physiology present, n (%) | 22 (4.4%) |
| Needle decompression performed, n (%) | 5 (1.0%) |
| Finger thoracostomy performed, n (%) | 40 (8.0%) |
| Technique with ICD insertion, n (%) | |
| Sterile | 189 (38.0%) |
| Emergent | 5 (1.0%) |
| Unknown | 302 (61.0%) |
| Prophylactic antibiotics given at insertion, n (%) | 95 (19.1%) |
| Post ICD Radiograph obtained, n (%) | 451 (90.7%) |
| Initial ICD malpositioned, n (%) | 51 (10.3%) |
| Corrective ICD procedures, n (%) | |
| ICD repositioning | 34 (6.8%) |
| ICD replacement | 13 (2.6%) |
LOS: Length of Stay; PHTX: Pneumothorax/Haemothorax; US: Ultrasound; CT: computed tomography; ICD: Intercostal Drain; SBP: systolic blood pressure
PEEP devices were provided to 40.8% of the cohort (n = 203), and less than a quarter of patients used an exercise bicycle (n = 103, 20.8%) or worked with physical therapy (n = 107, 21.5%) (Table 3). Most patients required no supplemental oxygen during their hospital stay (n = 463, 93.2%), and 2 patients enrolled in the protocol progressed to require non-invasive ventilation or intubation. Wall suction was initiated for 41 patients, or 8.2% of the cohort.
Table 3.
ICD care and complications
| Outcomes | All (n = 497) |
|---|---|
| Any PEEP device provided to patient, n (%) | 203 (40.8%) |
| Exercise bicycle used, n (%) | 103 (20.8%) |
| Physical therapy session, n (%) | 107 (21.5%) |
| Maximum oxygen needs during hospital stay, n (%) | |
| No supplemental oxygen | 463 (93.2%) |
| Nasal cannula, facemask or nonrebreather | 32 (6.4%) |
| Non-Invasive Ventilation | 1 (0.2%) |
| Intubation | 1 (0.2%) |
| Wall suction initiation, n (%) | 41 (8.2%) |
| Delayed complications, n (%)* | 38 (7.6%) |
| ICD dislodged requiring tube repositioning or replacement, n (%) | 9 (1.8%) |
| ICD insertion site superficial infection, n (%) | 9 (1.8%) |
| Empyema, n (%) | 3 (0.6%) |
| Non-ventilator pneumonia, n (%) | 5 (1.0%) |
| Retained haemothorax, n (%) | 6 (1.2%) |
| Respiratory failure requiring mechanical ventilation, n (%) | 1 (0.2%) |
| Post ICD removal reinsertion required, n (%) | 6 (1.2%) |
| Thoracic surgery, n (%) | 1 (0.2%) |
| Other# n (%) | 5 (1.0%) |
| Hospital length of stay in days, median (IQR) | 3.0 (2.0, 5.0) |
| Days with ICD inserted (from first ICD placed, to last ICD removed), median (IQR) | 3.0 (2.0, 4.0) |
| Survival to Hospital Discharge, n (%) | 497 (100.0%) |
| Readmission due to any cause within 30 days, n (%) | 14 (2.8%) |
* This total is listed as patient-level complications: patients may experience one or more of the delayed complications listed.
LOS: Length of Stay; PEEP: Positive End Expiratory Pressure; ICD: Intercostal Drain; IQR: Inter-Quartile Range
#Other complications: persistent hypoxia, gastropleural fistula, pleural effusion, residual pneumothorax, resolved pneumothorax but with persistent right lower lobe consolidation
Delayed ICD complications occurred for 38 patients (7.6% of cohort). The most common complications were infectious (n = 15) of which nine had a superficial ICD insertion site infection, three had empyema, and five had non-ventilator associated pneumonia. (Table 3) Of the three patients with empyema, one required intubation, two underwent surgical washout, and one was treated with antibiotics alone. Other delayed complications included retained haemothorax (n = 6, only one of which ultimately required thoracic surgery due to empyema), ICD dislodgment requiring ICD replacement (n = 9), and premature ICD removal by the clinician requiring replacement (n = 6). Hospital length of stay was a median of 3 days (IQR 2–4) and 51.6% of patients were discharged ≤ 72 h. The proportion with length-of-stay ≤ 72 h was non-significantly higher than the hypothesized proportion of 0.5 (Z = 0.76, p = 0.22), with a confidence interval that includes the null (95% CI 0.47–0.56). Length of stay is depicted in histogram format in Supplemental Fig. 1. All patients survived to hospital discharge, and only 14 patients (2.8%) were readmitted due to any cause within 30 days.
The following variables were significantly associated with prolonged length-of-stay (defined as > 72 h): haemothorax (OR 1.86), firearm injury (OR 11.18), presence of polytrauma (OR 3.1), receipt of blood products (OR 4.83), initial ICD placement at the tertiary hospital (OR 2.34), ICD management at the tertiary hospital (OR 4.15), wall suction initiation (OR 2.2), and presence of PHTX complication (OR 3.14) (Supplemental Table 1). The median NISS of patients that had ICD insertion at the tertiary hospital was 18 (IQR 9–25) compared to 10 (IQR 6–14) for patients with ICDs placed at district or regional facilities (p = < 0.01).
Discussion
The Western Cape ICD protocol appears effective for the treatment of simple traumatic pneumothorax and haemothorax, particularly with the use of careful patient selection criteria. Most patients enrolled in the protocol were young men in their 20s and 30s, injured by stabbing/cutting mechanisms, with isolated thoracic trauma. Over 90% of patients completed treatment without oxygen support or wall suction initiation, and the median hospital length of stay was 3 days.
The protocol appears to be safe, as complication rates were low. Our cohort had a composite complication rate of 16.7% at the patient-level, combining malpositioning and delayed complications. A systematic review by Hernandez et al. evaluating ICD complications in trauma care cited a median complication rate of 19%, with the caveat that historically, many studies did not follow a comprehensive approach to characterizing ICD complications [13]. This suggests that the Western Cape protocol is safe compared to standard of care.
Wall suction was utilized for less than 10% of the cohort, reserved for patients who had persistent air leak or haemothorax. This selective use of wall suction contrasts with practice patterns in other countries, where continuous wall suction is applied up front for all ICD patients [14]. While use of wall suction may be beneficial for accelerating recovery, our study suggests that it is not essential.
Our study parallels findings from Knottenbelt and van der Spuy’s studies conducted at Cape Town’s Groote Schuur hospital. Their 1990 study evaluated ICD drainage times among patients with traumatic pneumothorax and used similar early mobilization techniques [15]. Among 504 patients who underwent ICD placement, 66% underwent drain removal within 24 h and 20% required suction for air leak. Only 1.9% of patients who presented for follow-up required re-drainage, and there were no cases of empyema. The shorter length of stay in Knottenbelt’s study is likely related to exclusion of haemothorax cases, along with the protocols’ mandate for prompt removal of ICDs once the air leak has resolved. A 1994 study by these same authors focused on traumatic haemothorax, and among 1845 patients, average drainage time was 27.1 h, with 81.8% patients discharged within 48 h or less [16]. 15 patients (0.8%) developed empyema and 25 died (1.4%) with cause of death attributed to acute haemorrhage in most cases. These seminal studies, in conjunction with the present study, suggest that South African clinicians are facile at ICD management, and able to facilitate rapid discharge while maintaining a low infectious complication risk.
ICD insertion is an essential procedure in the management of chest trauma, forming a critical component of a physician’s procedural skill set, but one that is prone to complications [17]. In our study, approximately 10% of patients had malpositioned ICDs, and most cases were related to depth of insertion - either too deep or too shallow. Malpositioned ICDs are particularly common when the procedure is performed by junior doctors, often correlating with limited procedural experience [17–19]. While the emergency departments of the Western Cape hospitals are often staffed by junior clinicians, the level of training of the proceduralists was not captured in our study. South African clinicians likely gain rapid experience in ICD insertion given the large volumes seen in local trauma centres; this is evidenced by the fact that our study enrolled nearly 500 ICD patients across 4 hospitals in the span of one year. However, a prior study suggests that procedural competence is greater among practitioners at specialized trauma units, compared to operators at referral hospitals [20]. Regardless, increased procedural training outside of the clinical environment may allow for more systematic assurance of procedural safety. This can be accomplished through simulation, cadaver models, and formal assessment of procedural competence through checklists [21, 22]. Additionally, supervision by senior clinicians during ICD placement, standardized procedure notes, and audit and reporting mechanisms are likely to improve procedural success and patient safety [23, 24].
The sterility of the ICD procedures was not well-documented in our study, and prophylactic antibiotics were given in less than 20% of cases. Despite the low rate of prophylactic antibiotics, our study found a less than 1% incidence of empyema, similar to rates from other South African studies [15, 16], as we have previously alluded to, and lower than previously published rates in other countries [25, 26]. A 2019 meta-analysis by Ayoub et al. found that use of prophylactic antibiotics with ICD insertion reduced the risk of empyema (RR 0.25) and pneumonia (RR 0.41) compared to placebo [25]. The Eastern Association for the Surgery of Trauma (EAST) conducted a similar analysis in 2022 and found that patients who received antibiotic prophylaxis were less likely to develop empyema (OR 0.47) but found no significant effects for pneumonia incidence or mortality [26]. EAST therefore conditionally recommends antibiotic prophylaxis with ICD placement. More studies are needed to determine the utility of prophylactic antibiotics.
The implementation of the protocol’s exercise prescription was highly variable. Less than half the patients were given PEEP devices, and less than a quarter used an exercise bicycle or worked with physical therapy. This suggests that many patients in the ICD protocol relied exclusively on ambulation for their pulmonary rehabilitation. Physiotherapy is an integral part of recovery from truncal trauma, although there is limited high-quality research about specific intervention bundles [27]. An international survey of physiotherapists found that over 90% relied on deep breathing exercises, active cough, and mobilization when working with patients who have sustained chest trauma. The rationale for these manoeuvres includes improvement of tidal volumes, enhanced ventilation/perfusion matching, and facilitation of secretion clearance to accelerate return to normal pulmonary function [28]. Two historical studies performed in South Africa have found that immediate physiotherapy after ICD placement was associated with shorter duration to haemothorax resolution and shorter hospital length of stay [29, 30]. Deep breathing with incentive spirometry (PEEP devices) also appears to be effective in lowering patients’ perception of pain [31]. Further research is required regarding the benefits of mobilization for PHTX care.
Bivariate logistic regression identified firearm injuries, presence of polytrauma, receipt of blood products, wall suction initiation, care at tertiary facility, and presence of ICD complications as risk factors for prolonged length of stay beyond 3 hospital days. This model suggests that patients with high impact mechanisms, haemorrhage, or multicompartment injuries are not good candidates for this protocol. Stringent patient selection is therefore imperative.
The safety and effectiveness of the Western Cape ICD protocol suggests that it may be applicable in other settings with high volumes of traumatic PHTX and limited access to resources such as wall suction or inpatient beds. Other Sub-Saharan African countries also see large volumes of penetrating traumatic chest injury and could be candidates for adopting the Western Cape ICD protocol. For example, studies from Nigeria and Senegal have found nearly half the cases of thoracic trauma were due to penetrating mechanisms related to interpersonal violence [32, 33]. However, we acknowledge the safety of this protocol in the Western Cape may partly be due to contextually-specific factors such as the extremely high volumes of this procedure in this health system. Future research is needed to verify the safety of this protocol in new contexts.
This study has several limitations. First, the study relied on data abstraction from the written medical record. The clinical notes regarding ICD insertion were brief, and frequently omitted information regarding sterile technique. This potentially reflects the limited time available for detailed procedural recording among doctors working in an overburdened health system. Additionally, nursing documentation did not always mention provision of PEEP devices, or document patient mobilization. As a result, such interventions may have gone undocumented despite having been performed. A study with direct observation of procedures and interventions may offer more accurate insights compared to retrospective reviews of clinical notes; however, such a study would likely be biased by the Hawthorne effect. Finally, this study did not include a control group. The Western Cape ICD protocol was implemented by all study sites, as it is considered the standard practice in this region. Therefore, the data presented are limited to descriptive measures.
In conclusion, this study found that for patients with simple traumatic PHTX, a protocol that involves ICD insertion using a dry seal valve, accompanied with early mobilization within an ED observation unit, is safe and effective, with median discharge within 3 hospital days and complication rates that are comparable to the standard of care.
Supplementary Information
Below is the link to the electronic supplementary material.
Supplementary Material 1: Supplemental Table 1. Bivariate logistic regression to evaluate factors associated with prolonged length-of-stay (>72 hours) among patients treated by the Western Cape ICD Protocol
Supplementary Material 2: Supplemental Fig. 1. Histogram showing hospital length of stay for patients treated using the Western Cape ICD protocol
Acknowledgements
The authors are grateful to the leadership, practitioners, and staff of the Western Cape Government Department of Health and Wellness for their long-standing commitment to trauma research and partnership with the EpiC study. This research would not be possible without the dedicated efforts of the Collaborative for Emergency Care in Africa (CECA), including Dr. Mandy Abels, Ms. Razia Amien, Dr. Kirstin Becker, Dr. Saeeda Benjamin, Dr. Raeesah Bray, Mr. Anthony Carr, Dr. Melany Codling, Ms. Zuleiga Davids, Ms. Magzane Ghall, Ms. Daisy Inglis, Ms. Alana Jooste, Mr. Vusumzi Kose, Dr. Mudiwa Llobell, Dr. Jani Marais, Sr. Minnie Minnaar, Ms. Tanisha Morris, Dr. Anna Notten, Dr. Azia Parker, Mr. Justin Petersen, Dr. Tahrir Rasool, Dr. Astrid Samuels, Dr. Kirsten Scoble, Ms. Regan Smith, Dr. Jarryd Thompson, Dr. Alexa van der Walt, Mr. Dylan van Staden, and Ms. Catherine Wandie.
Group Authors:
“EpiC Study Site Collaborators” author group is comprised of: Mohammed Mayet, MBChB – Khayelitsha Hospital; Lesley Hodsdon, MBChB – Worcester Regional Hospital; L’Oreal Snyders MBChB – Delft Community Health Centre; Leigh Wagner MBChB – Khayelitsha Site-B Community Health Centre; Karlien Doubell MBChB – Ceres District Hospital; Denise Lourens, MBChB – Worcester Forensic and Pathology Services Laboratory. All are employees of Western Cape Government Department of Health and Wellness, Cape Town, South Africa.
Abbreviations
- ICD
Intercostal drains
- PHTX
Pneumothorax or haemothorax
- PEEP
Positive end-expiratory pressure
- EpiC
Epidemiology and Outcomes of Prolonged Trauma Care study
- ICU
Intensive care unit
- REDCap
Research Electronic Data Capture
- IQR
Inter-Quartile Range
- LOS
Length of stay
- HIV/AIDS
Human Immunodeficiency Virus/Acquired Immunodeficiency Syndrome
- AIS
Abbreviated Injury Scale
- NISS
New Injury Severity Score
- SBP
Systolic blood pressure
- SpO2
Blood oxygen saturation
- TXA
Tranexamic acid
- US
Ultrasound
- CT
Computed tomography
- CI
Confidence Interval
- OR
Odds Ratio
- RR
Relative Risk
- EAST
Eastern Association for the Surgery of Trauma
- ED
Emergency Department
Author contributions
SB, JD, and NM conceptualized and developed the research question. SB, JD, SS, JM, and NM obtained funding for the study. NM, HL, GO, SdV, and ESSC provided oversight of study activities. LF and MR curated the data and created the visualizations and LF performed the analysis. JW supervised the analysis. SB and RA drafted the manuscript. All authors were involved in reviewing and editing the manuscript, and all authors read and approved the final version.
Funding
This research was funded through grants from the U.S. Defense Health Agency (contract numbers W81XWH-20-2-0042 and HT9425-24-C-0123).
Data availability
The study data is not publicly available due to funding agency requirements; however, they can be provided by the corresponding author upon reasonable request, and upon satisfying contractual and regulatory requirements.
Declarations
Ethics approval and consent to participate
This study received ethics approval from Stellenbosch University Health Research Ethics Committee (Reference N20/03/036) under a waiver of consent, as this is considered standard of care. The Defense Health Agency’s Office of Human Research Oversight, the funder’s ethics committee, provided an additional review and determined compliance with the U.S. Department of Defense’s human subject research protections (OHRO Log Number E01863.1x).
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Beyer CA, Ruf AC, Alshawi AB, Cannon JW. Management of traumatic pneumothorax and hemothorax. Curr Probl Surg. 2025;63:101707. [DOI] [PubMed] [Google Scholar]
- 2.Blank JJ, de Moya MA. Traumatic pneumothorax and hemothorax: What you need to know. J Trauma Acute Care Surg. 9900. [DOI] [PubMed]
- 3.Anderson D, Chen SA, Godoy LA, Brown LM, Cooke DT. Comprehensive review of chest tube management: a review. JAMA Surg. 2022;157(3):269–74. [DOI] [PubMed] [Google Scholar]
- 4.Brown SG, Ball EL, Perrin K, Asha SE, Braithwaite I, Egerton-Warburton D, et al. Conservative versus interventional treatment for spontaneous pneumothorax. N Engl J Med. 2020;382(5):405–15. [DOI] [PubMed] [Google Scholar]
- 5.Brims FJH, Maskell NA. Ambulatory treatment in the management of pneumothorax: a systematic review of the literature. Thorax. 2013;68(7):664. [DOI] [PubMed] [Google Scholar]
- 6.Hallifax RJ, McKeown E, Sivakumar P, Fairbairn I, Peter C, Leitch A, et al. Ambulatory management of primary spontaneous pneumothorax: an open-label, randomised controlled trial. Lancet. 2020;396(10243):39–49. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Walker SP, Keenan E, Bintcliffe O, Stanton AE, Roberts M, Pepperell J, et al. Ambulatory management of secondary spontaneous pneumothorax: a randomised controlled trial. Eur Respir J. 2021;57(6):2003375. [DOI] [PubMed] [Google Scholar]
- 8.Kithuka CM, Ntola VC, Sibanda W, Afr. J Surg. 2023;61(3):12–6. [PubMed] [Google Scholar]
- 9.Clarke DL, Quazi MA, Reddy K, Thomson SR. Emergency operation for penetrating thoracic trauma in a metropolitan surgical service in South Africa. J Thorac Cardiovasc Surg. 2011;142(3):563–8. [DOI] [PubMed] [Google Scholar]
- 10.Bhana M, Fru P, Plani F. A long walk to freedom: the epidemiology of penetrating trauma in South Africa–analysis of 4 697 patients over a six-year period at Chris Hani Baragwanath Academic Hospital. S Afr J Surg. 2022;60(2):77–83. [PubMed] [Google Scholar]
- 11.Cooper C, Hardcastle T. Xpand chest drain: assessing equivalence to current standard therapy - a randomised controlled trial. S Afr J Surg. 2006;44(4):132–5. [PubMed] [Google Scholar]
- 12.Suresh K, Dixon JM, Patel C, Beaty B, del Junco DJ, de Vries S, et al. The epidemiology and outcomes of prolonged trauma care (EpiC) study: methodology of a prospective multicenter observational study in the Western Cape of South Africa. Scand J Trauma Resusc Emerg Med. 2022;30(1):55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.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(2):410–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Feenstra TM, Dickhoff C, Deunk J. Systematic review and meta-analysis of tube thoracostomy following traumatic chest injury; suction versus water seal. Eur J Trauma Emerg Surg. 2018;44(6):819–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Knottenbelt JD, van der Spuy JW. Traumatic pneumothorax: a scheme for rapid patient turnover. Injury. 1990;21(2):77–80. [DOI] [PubMed] [Google Scholar]
- 16.Knottenbelt JD, Van der Spuy JW. Traumatic haemothorax–experience of a protocol for rapid turnover in 1,845 cases. S Afr J Surg. 1994;32(1):5–8. [PubMed] [Google Scholar]
- 17.Corcoran JP, Hallifax RJ, Talwar A, Psallidas I, Sykes A, Rahman NM. Intercostal chest drain insertion by general physicians: attitudes, experience and implications for training, service and patient safety. Postgrad Med J. 2015;91(1075):244–50. [DOI] [PubMed] [Google Scholar]
- 18.Kong V, Oosthuizen G, Sartorius B, Keene C, Clarke D. An audit of the complications of intercostal chest drain insertion in a high volume trauma service in South Africa. Annals Royal Coll Surg Engl. 2014;96(8):609–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Platnick C, Witt CE, Pieracci FM, Robinson CK, Lawless R, Burlew CC, et al. Beyond the tube: Can we reduce chest tube complications in trauma patients? Am J Surg. 2021;222(5):1023–8. [DOI] [PubMed] [Google Scholar]
- 20.Maritz D, Wallis L, Hardcastle T. Complications of tube thoracostomy for chest trauma. S Afr Med J. 2009;99(2):114–7. [PubMed] [Google Scholar]
- 21.Tan TX, Buchanan P, Quattromani E. Teaching residents chest tubes: simulation task trainer or cadaver model? Emerg Med Int. 2018;2018(1):9179042. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Salamonsen MR, Bashirzadeh F, Ritchie AJ, Ward HE, Fielding DIK. A new instrument to assess physician skill at chest tube insertion: the TUBE-iCOMPT. Thorax. 2015;70(2):186. [DOI] [PubMed] [Google Scholar]
- 23.Ball CG, Lord J, Laupland KB, Gmora S, Mulloy RH, Ng AK, et al. Chest tube complications: how well are we training our residents? Can J Surg. 2007;50(6):450–8. [PMC free article] [PubMed] [Google Scholar]
- 24.Collinge CA, McWilliam-Ross K, Kelly KC, Dombroski D. Substantial improvement in prophylactic antibiotic administration for open fracture patients: results of a performance improvement program. J Orthop Trauma. 2014;28(11):620–5. [DOI] [PubMed] [Google Scholar]
- 25.Ayoub F, Quirke M, Frith D. Use of prophylactic antibiotic in preventing complications for blunt and penetrating chest trauma requiring chest drain insertion: a systematic review and meta-analysis. Trauma Surg Acute Care Open. 2019;4(1):e000246. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Freeman JJ, Asfaw SH, Vatsaas CJ, Yorkgitis BK, Haines KL, Burns JB, et al. Antibiotic prophylaxis for tube thoracostomy placement in trauma: a practice management guideline from the Eastern Association for the Surgery of Trauma. Trauma Surg Acute Care Open. 2022;7(1). [DOI] [PMC free article] [PubMed]
- 27.van Aswegen H. Physiotherapy management of patients with trunk trauma: A state-of-the-art review. South Afr J Physiotherapy. 2020;76(1):1406. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Aswegen Hv, Reeve J, Beach L, Parker R, Olsèn MF. Physiotherapy management of patients with major chest trauma: results from a global survey. Trauma. 2020;22(2):133–41. [Google Scholar]
- 29.Senekal M, Eales C. The optimal physiotherapeutic approach to penetrating stab wounds of the chest. South Afr J Physiotherapy. 1994;50(2):29–36. [Google Scholar]
- 30.Ngubane T, de Charmoy S, Eales C. The treatment of stabbed chests at Ngwelezana hospital, KwaZulu-Natal. South Afr J Physiotherapy. 1999;55(1):23–5. [Google Scholar]
- 31.Kourouche S, Buckley T, Munroe B, Curtis K. Development of a blunt chest injury care bundle: An integrative review. Injury. 2018;49(6):1008–23. [DOI] [PubMed] [Google Scholar]
- 32.Adal O, Tareke AA, Bogale EK, Anagaw TF, Tiruneh MG, Fenta ET, et al. Mortality of traumatic chest injury and its predictors across sub-saharan Africa: systematic review and meta-analysis, 2024. BMC Emerg Med. 2024;24(1):32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Okugbo S, Okoro E, Irhibogbe P. Chest trauma in a regional tRAUMA centRE. J West Afr Coll Surg. 2012;2(2):91–101. [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Material 1: Supplemental Table 1. Bivariate logistic regression to evaluate factors associated with prolonged length-of-stay (>72 hours) among patients treated by the Western Cape ICD Protocol
Supplementary Material 2: Supplemental Fig. 1. Histogram showing hospital length of stay for patients treated using the Western Cape ICD protocol
Data Availability Statement
The study data is not publicly available due to funding agency requirements; however, they can be provided by the corresponding author upon reasonable request, and upon satisfying contractual and regulatory requirements.