This study aimed to describe ETI in South Africa in terms of patient demographics, indications for intubation, means of intubation and success rates. To our knowledge, this is the largest study of paramedic-performed ETI from the African continent and other low-resource settings. We found that most patients who underwent ETI during this period were adults, males, trauma victims, or had a decreased level of consciousness following trauma. Non-physician ETI appeared to have high overall success rates, despite the presence of risk factors for difficult intubation. The most common approach to ETI was RSI.
South Africa, like many other LMICs, has a tremendously high trauma burden. 20,21 It is therefore not surprising that the predominant reason for emergency care was following injury. Injury, and particularly TBI, is one of the most important contributors to morbidity and mortality in LMICs, especially in the younger, economically active younger population. 22 Out-of-hospital cardiac arrest was also a common presentation and this is likely reflective of an increasing burden of cardiovascular disease in Sub-Saharan Africa, including South Africa. 23 ETI in the setting of cardiac arrest is only recommended under optimal conditions and in settings with demonstrable high success rates, but has further been de-emphasised with chest compressions as the priority. 24 The utility of ETI in the South African context, where out-of-hospital cardiac arrest survival rates are very low, 25,26 is yet to be determined.
Across the world, non-physician ETI FPS rates range from 47–98%. 27–29 Further, a recent systematic review that was limited to ETI with an RSI approach only, found non-physician FPS of 78% (95% CI [65%-89%]). 2 The FPS rates reported in this study (75%) compares favourably to that reported in the international literature, despite comprising ETI approaches other than RSI. The use of neuromuscular blocking agents have been found to decrease the risk of difficult intubation, 30 and yield a higher FPS rate. 31 This was also demonstrated in our study where RSI was shown to improve the odds of FPS over other approaches. Consequently, these other approaches could have had a lowering effect on the reported FPS. A third of cases had to be excluded because the number of intubation attempts was not recorded. It is not possible to know whether this was more likely to be noted when one or multiple intubation attempts were made. It is therefore conceivable that this could have influenced the reported FPS rate. Similarly, the overall success rates (95.7%) in this sample also compared favourably to that reported elsewhere (97% (95% CI [95–99%]) in RSI only. 2 When comparing these rates with non-physician ETI using multiple approaches, the overall success rate is slightly higher in our study than reported in a recent meta-analysis (91.7 (95% CI 61.6–100)). 3
Owing to heterogeneity in prehospital emergency medical services across the world in terms of provider profile and skill level, and resourcing, comparisons of FPS is not always appropriate, and this should be taken into consideration when interpreting the results.
Perhaps then, it might be more appropriate to compare our FPS rates in this study with other studies originating from South Africa. A recent retrospective descriptive analysis of ETIs in Helicopter Emergency Medical Services (HEMS) reported a FPS rate of 79%, and an overall success rate of 98%. 14 This study included all approaches to ETI, and again compares favourably to the success rates reported herein for ground-based EMS. In another study, prehospital emergency care students achieved FPS and overall success rates of 85.2% and 92.4% when using an RSI approach only. 13 This is a higher FPS rate than reported in our study, but this could again be explained by the utility of neuromuscular blocking agents. Lastly, when comparing prehospital with emergency department success rates in South Africa, a recent study reported an FPS of 81.8%, which is considerably higher than reported herein. However, a sub-analysis of this sample reveals an FPS rate of 73.3% in cases where direct laryngoscopy was attempted, versus video laryngoscopy. 32 Video laryngoscopy was not available in the ground-based EMS involved in this study and thus, the latter FPS is a more appropriate comparison. Another consideration when comparing the FPS rates of this study with our results is the low proportion of trauma patients (20.9%). Manual in-line neck stabilisation, common practice during ETI in trauma victims, has been shown to significantly increase difficulty and failure of ETI. 33
Following multiple regression analysis, three factors remained associated with FPS: RSI approach, ventilation as an indication for intubation, and on scene time. RSI was associated with overall success, and inversely related to failed intubation. The impact of an RSI approach on intubation difficulty and success rates has already been discussed.
On scene time was inversely associated with FPS, while increasing on scene time was associated with overall failure. It is a logical conclusion that the requirement for multiple intubation attempts may prolong scene time, while successfully securing the airway on first attempt will limit the time spent on scene for stabilisation. This was demonstrated in a South African HEMS-based study where the number of clinical interventions were correlated with scene, and every 1 additional intervention increased scene time by approximately 4 minutes. 1 Importantly though, interventions (with ETI being one of the most prevalent) did not result in a significantly more stable patient. The effect of prolonged scene time on mortality is yet to be determined in the South African context, especially with such a high burden of injury.
Ineffective ventilation as indication for ETI was inversely associated with FPS. This might be explained by the predictable instability associated with acidosis and hypoxaemia that accompany hypoventilation. 34 This may preclude prolonged attempts at securing the airway and result in earlier termination of an intubation attempt to avoid adverse events.
While success rates are a useful measure of airway management, they can be misleading as surrogates for safe ETI. Instead, there is a drive towards reporting of peri-intubation adverse events, rather than simply relying on success rates. In this retrospective study, it was difficult to validly extract AEs from the PRFs as the exact time of intubation was not recorded in most instances. Poor reporting was also the reason why a third of eligible cases where the number of intubation attempts were not recorded, had to be excluded. We therefore suggest that PRFs and/or reporting documents for all PHAAM instances be adjusted to allow for meaningful analysis as part of quality improvement and research practices. Using standardised airway forms has been found to reduce the rate of missing information and significantly increase the quality of data reported during PHAAM. 35 Calls for standardisation and robust clinical governance for PHAAM in South Africa, have been made previously 12 but there seem to be barriers to their implementation. 36
During the data collection period of this study, all advanced life support prehospital providers were licensed to intubate patients. In 2020, the Health Professions Council of South Africa (HPCSA) made the decision to remove ETI from the scope of practice of all prehospital providers and subsequently prehospital intubation can only be performed by degreed providers using the RSI approach. Currently, there are only 900 degreed providers registered with the HPCSA, 37 yet it is unclear how many are actually still in full-time clinical practice in South Africa - a major concern as South Africa has had some considerable challenges in retention of prehospital providers. 38,39
Emergency intubation in the prehospital environment is a complex intervention with severe complications if it is poorly planned or performed, so optimising all factors involved prior to intubation makes sense. This would include allowing only well-trained, competent individuals with adequate skills and experience to undertake PHAAM. 40,41 While it can be seen as a commendable decision to reserve intubation for the highest qualified prehospital providers, this might translate into lack of access to a potentially life-saving intervention early in the course of emergency care. This may be undesirable, especially in TBI where early control of oxygenation and ventilation may prevent secondary brain injury 42 - TBI comprised over two-thirds of the patients in this study. The potential impact and unintended consequences of removing access to ETI in these patients warrants urgent examination. Some solutions to this may be to develop retention strategies for degreed paramedics and incentivise them to remain in practice, especially in rural or underserved communities. Another option may be regionalised scopes of practice for all ALS that are tailored to anticipated prehospital times and local injury and illness epidemiology.
It is essential to acknowledge the paucity of robust data on the effect of prehospital ETI on morbidity and mortality, 43 especially in trauma. 6,41,44 Where data exists, it is mostly from high income settings, or cannot allow for meaningful comparison or meta-analysis owing to health system variation, selective reporting, or risk of bias. There is an urgent need to performance additional research that evaluates the peri-intubation safety and outcome following PHAAM. In our view, the only way that this could happen robustly is through the implementation of mandatory standard reporting databases.
This retrospective study is not without limitations. Data were extracted from self-reported clinical notes that are not intended for research and therefore had to be extracted based on the clinical impression of the extractors. Certain data such as airway difficulty and aggravating factors relating to airway assessment are based on the subjective judgement of the practitioner performing the assessment and can therefore only be considered as estimates of potential airway difficulty. External validity is certainly affected by the inclusion of private services and only one, provincial public emergency medical service.