Prevalence and Factors Associated with Mortality among Chest Injury Patients Admitted at Muhimbili National Hospital in Dar es Salaam, Tanzania

Introduction: Accidents represent a significant proportional of non-com-municable disease in the current century, and chest injury is common. However, management and outcome of these injuries is poor in low resource setting like Tanzania. The aim of this study was to determine the prevalence and factors associated with mortality among chest injury patients at a tertiary level health facility in Tanzania. Method: A prospective Cross-Sectional study of chest injuries among trauma patients attended at Muhimbili National Hospital between September 2019 and February 2020. Results: A total of 282 trauma patients were seen, out of which 51/282 (18.1%) sustained chest injury. Road Traffic Crashes were the leading cause of chest injury 41/51 (80.4%). Majority 17/51 (33.3%) presented with lung contusion, followed by pneumohemothorax and rib fractures each 8/51 (15.7%). Most of the patients 27/51 (52.9%) were managed by tube thoracostomy and 42.1% conservatively. Mortality was 11/51 (21.6%). Independent factors associated with mortality were: Associated injuries (Odds Ratio (OR) 0.07, 95% CI 0.01


Introduction
Trauma continues to be associated with high morbidity and mortality both in developed and developing countries [1]. Chest Injuries account for 10% of Global trauma admission and 25% of trauma related deaths [2] [3]. Studies have revealed the prevalence of chest injury to be varying from different parts of the world, being high in low-and middle-income countries and low in developed countries [4], this is mostly due to variations in preventive measures of trauma.
The estimated mortality due to chest injuries in Tanzania is 40% [5], and continues to be the commonest cause of surgical admissions with significantly high morbidity and mortality [6].
Causes and pattern of chest injuries vary in different parts of the world due to socio-economic status variations, the commonest cause being road traffic crashes [7] [8], majorities of victims being in the active age group sustaining blunt chest injury. Like other developing nations in the world, Tanzania has a significantly high rate of traffic related deaths and disabilities. A hospital-based injury surveillance [9] revealed traffic crashes to be the leading cause of injuries accounting for 47.5% of all injuries seen and 60.5% of injuries mortality. Chest injury is second only to head injury as a major cause of morbidity and mortality in Tanzania emergency rooms, and this can be explained by the lack of organized pre-hospital care, severity of injuries and late management of patients [10].
Management option depends on type of chest injury and clinical presentation of the patients. Patients with pneumothorax, haemothorax or both would improve on tube thoracostomy. Other patients would require mechanical ventilation, appropriate analgesics management, supportive therapy and critical care observation [11] [12]. It is therefore necessary for accurate, early identification and aggressive management of chest injuries, along with prompt treatment of associated injuries for optimal patient outcome. This study was therefore conducted to help us understand the magnitude and management of chest injury patients at a tertiary level health facility in Tanzania. The result of this study will help in establishing prevention strategies as well as management protocol to better assess, treat and monitor chest injury patients with a view of improving patient outcomes. Data collection and management: a structured questionnaire designed by principal investigators was used to collect information from the study participants in a face-to-face interview, and from medical records. Questions were drawn from previously conducted studies and a pilot study was conducted to ensure validity and reliability of the data collection tool. A questionnaire had 18 items including: socio-demographic characteristics, mode of injury, type of injury, severity of injury, treatment pattern and patient's outcome (Appendix). Statistical analysis: all variables were categorized and described using frequency distribution. Chi-square test was used for bivariate analysis and those variables with observed frequency less than five Fisher's exact test was applied. A variable with (p ≤ 0.05) with mortality was considered to be statistically significant. Variables that demonstrated significant bivariate association with mortality were entered into the multivariate logistic regression modal to assess independent effects. Parameter of measurement to assess association was odds ratio.

Methodology
Ethical consideration: ethical approval for the study was obtained from the Muhimbili University of Health and Allied Sciences Research Ethics Committee.
An informed written consent was sought from patients or relatives.

Results
A total of 282 trauma patients were seen between September 2019 and February 2020. Chest injuries accounted for 51/282 (18.1%) of patients, majority (72.5%) belonged to the productive age group (20 to 39 years), and 74.5% were males, with a male to female ratio of 3:1 (Table 1).

Mechanism of Injury
Road traffic crashes (80.4%) were responsible for the majority of Chest injuries, Assault (15.7%) and fall from height (3.9%).

Type of Chest Injuries
Blunt trauma accounted for 88.2% of the chest injuries, 11.8% were penetrating injuries. Lung contusion was the commonest (33.3%) followed by, Pneumohemothorax and Rib fractures each (15.7%). Pneumothorax was present in 11.7% of the patients whereas haemothorax was present in 7.8% of the cases. International Journal of Clinical Medicine Two patients (3.9%) had an injury to the heart and 3.9% had a Flail chest ( Table   2). Majority of patients (52.9%) had tube thoracostomy done, and 42.1% managed conservatively.   The following factors remained statistically significant after multivariate logistic regression analysis: Associated injuries, critical injury, analgesia mode, time to treatment, bilateral chest involvement and invasive ventilation. Those with no associated injuries were 93% less likely to die than those who had associated injuries. Critical injured patients were 24 times more likely to die than moderately injured patients. Presenting to hospital more than 24 hours post-injury were 5.5

Factors Associated with Mortality
times more likely to die compared to those who presented to hospital within 24 hours post-injury. Those with bilateral chest involvement were 4.6 times more likely to die than those who had unilateral chest involvement. The risk of death was 31.5 times greater in those who needed invasive ventilation than non-invasive ventilation, and it was 78% less in those who had more than one analgesia than those managed by a single analgesia (Table 3).

Discussion
This study aimed at finding the prevalence of chest injuries among trauma patients and the predictors of mortality.  [20], in which they reported few patients with life threatening conditions such as oesophageal perforation, cardiac injury and diaphragmatic rupture. Majority of victims sustaining these types of severe injuries die at the site of accidents and hence don't make it to hospitals. Majorities 58.8% of injury victims were attended in within 24 hours from injury in this study, and were more likely to survive than those who received medical attention more than 24 hours from injury. Delay in receiving medical attention minimises the chance of survival for the injured patients, especially severely injured patients. This calls for proper pre-hospital emergency care system and ambulance services in developing countries like Tanzania if severely injured patients are to survive.
Most of the patients (52.9%) in this study, required a chest tube thoracostomy and others (42.1%) were managed conservatively by (observation, analgesia and antibiotics and chest physiotherapy), similar to what has been reported in other studies done elsewhere [6] [23]. It is also reported in studies done in developed countries that observation, chest tube placement, adequate volume replacement, occasional respiratory support and serial chest X-rays are the only treatment required in 80% -85% of the patients [24] [25].
In this study, 23 (45%) patients were given one type of pain medication and 28 (55%) were given more than one type of pain medication. Pain management option significantly influenced survival, those who received more than 1 type of pain medication were more likely to survive as compared to those who received a single type of pain medication. Results similar to Annalise et al. in her review on treatment of chest trauma and their impact [26], in which it was revealed that using more than one option of pain medication improved the outcome of thoracic injury patients. Other studies have demonstrated a big role of regional anaesthesia with intercostal blockade, thoracic epidurals and paravertebral blockades in significant reduction of pain in chest injury patients [6] [27]. The main options in our setting were systemic medication (Opioids, NSAIDS and paracetamol), which have been reported in literatures to be insufficient for optimal pain control [28]. Optimal pain control in chest injury patients prevents splinting of the diaphragm and atelectasis.
In this study, 12 (23.5%) patients needed invasive ventilation and were 31.5 times more likely to die than non-invasive ventilated patients. Patients with blunt or penetrating chest trauma may require mechanical ventilation. Lung protective ventilation strategies have to be applied otherwise it is associated with attributable mortality if it is set incorrectly [29], hence a need for mechanical ventilation modes training for personnel's working in our emergency rooms and intensive care units.
The  [20]. There is therefore a need for management improvement strategies targeting these predictors of mortality in order to improve outcome of chest injury patients in Tanzania, including establishing trauma care system in the country.
In considering the findings of this study it is important to bear in mind the following limitations: firstly, this was a single centre study with small number of patients, and the time frame of the study was short, hence it may not reflect what is happening in other centres. Secondly, data collectors not collecting all data and so some are missing. Thirdly, information bias from participants and data collectors may have affected the quality of data.

Conclusion
The results of this study provide valuable insight into the burden and manage-

What Is Known about This Topic?
• Accidents represent a significant proportional of non-communicable disease in Tanzania. • Majority are due to road traffic crashes.
• Chest injury is among the common injuries sustained by the victims.

What This Study Adds
• This study provides valuable insight into the burden and management of Chest injury in Tanzania hospitals.
• It also points to the need of establishing management guidelines for chest trauma in both pre-hospital and in-hospital settings.