The management of an elderly patient is significantly different from that of a young patient due to the physiological and pharmacological changes associated with advanced age, with an increased risk of complications [1] [2] . The World Health Organisation (WHO) defines an elderly person as an individual aged 60 or over. According to the WHO, by 2050, the world’s population aged 60 and over is expected to reach 2 billion, compared with 900 million in 2015 [3] . However, in the National Institute of Aging distinguishes between three categories of older people older people: the “young old” (65-74), the “old old” (75 - 84) and the “oldest old” (75 - 84) and the “very old” (85 and over) [4] . We need to distinguish between different profiles of elderly patients, for whom risk assessment and perioperative needs will be quite different [1] : The vigorous or robust: they are autonomous with few associated co-morbidities and only present physiological and pharmacological changes related to their age. Fragile patients frail: they are at higher risk of mortality, sequelae or progression to dependency in the event of an intercurrent event (confusion, fall, infection, etc.). However, screening for frailty, combined with active physical, cognitive, nutritional and social care, can limit the consequences of this frailty on postoperative outcomes. Polypathological dependents: the identification of co-morbidities and addictions often plays a greater role in risk assessment than age itself, and enables these risks to be weighed against the expected benefits of the proposed intervention.

The ageing of the world’s population represents a major challenge for healthcare systems, including in low- and middle-income countries (LMICs). According to the United Nations, the proportion of people aged 60 and over will increase significantly over the next few decades, including in sub-Saharan Africa and other regions with limited resources [5] . This phenomenon is accompanied by a rise in degenerative musculoskeletal conditions, such as osteoarthritis and osteoporosis, leading to an increase in demand for orthopaedic surgery, particularly for femoral neck fractures and joint replacements [6] [7] .

Orthopaedic surgery in the elderly is particularly risky in terms of anaesthesia. This population often has multiple co-morbidities (heart disease, diabetes, renal failure, etc.), reduced physiological reserves and an altered response to surgical stress [8] . These factors, combined with frailty and multiple medications, increase the risk of intraoperative and postoperative complications such as cardiovascular events, delirium, infections and mortality [9] [10] .

In high-income countries, multidisciplinary protocols have been introduced to improve perioperative management of the elderly. Appropriate anaesthetic techniques, rigorous preoperative optimisation and continuous postoperative monitoring have significantly reduced morbidity and mortality in this population [11] . However, in low- and middle-income countries (LMICs), anaesthetic management is still subject to numerous constraints: shortage of qualified staff, lack of monitoring equipment, delays in surgical management, and lack of protocols adapted to this age group [12] [13] . These difficulties are all the more worrying given that elderly patients are particularly vulnerable, with in-hospital mortality rates of up to 20% following femoral neck fractures in certain African contexts [14] .

Meta-analyses have also demonstrated a reduction in mortality, postoperative complications and hospital length of stay when a protocol of intraoperative haemodynamic control is applied to medium- and high-risk subjects undergoing major surgery [15] [16] . In Turrentine’s cohort [15] , 25% of patients were aged over 70, and the number of risk factors increased with age. Aubrun [1] found that age was an independent factor in postoperative morbidity and mortality, with a four-fold increase in risk after the age of 75 for surgery on the upper end of the femur. A Brazilian study of 182 patients undergoing femoral fracture surgery found a role for age, high ASA class and a history of pulmonary embolism in morbidity and mortality [17] . In Africa, studies on anaesthesia for orthopaedic surgery are rare. Chaibou [18] in Niamey found a mortality rate of 1.68% in 119 elderly patients undergoing all types of surgery. Mbombo’s recently published study on anaesthesia in the elderly, taking all pathologies together, found that age was not an independent factor in mortality (which was 5.9%), but rather high ASA class, the severity of the surgical attack and anaesthesia duration of more than two hours [19] . A study conducted in Tanzania showed a post-operative complication rate of 36% in patients over 65 undergoing major orthopaedic surgery [20] .

In Mbombo’s study cited above [19] in Kinshasa on anaesthesia for elderly patients, orthopaedics ranked third behind general and digestive surgery and urology. However, several studies suggest that a small proportion of post-operative mortality is the direct consequence of the pathology that led to the surgical procedure, particularly in the case of femoral neck fractures, where 24% of deaths are linked to the fracture. Despite the importance of the subject, few studies have been devoted to assessing the factors associated with morbidity and mortality in anaesthesia for elderly patients in countries with limited resources. Most of the available data come from high-income countries, and do not take account of specific local realities. It is therefore essential to produce contextualised data in order to gain a better understanding of the determinants of complications and to suggest ways of improving them that are adapted to local contexts.

This study is part of that effort. It aims to identify the factors associated with morbidity and mortality during anaesthesia for orthopaedic surgery in elderly patients in a hospital in a low-income country. By analysing demographic, clinical and anaesthetic characteristics, this study aims to make a contribution to the local and international scientific literature, by highlighting relevant levers for action to optimise the management of this high-risk population.

This study was conducted to determine the morbidity and mortality encountered during anaesthesia for orthopaedic surgery in the elderly and the factors associated with it. This type of anaesthesia concerns 1.2% of all anaesthetised patients, 18.67% of all elderly patients and 0.8% of all orthopaedic surgery patients. The average age is 68.5, with a slight male predominance, and the surgical procedures are often major. Spinal anaesthesia is the most commonly used anaesthetic technique and arterial hypotension is the most frequent intraoperative complication, while intraoperative transfusion accounts for 20%. Pulmonary embolism (1.5%) was the complication responsible for two deaths. The presence of cardiovascular co-morbidity or thromboembolic risk, major surgery and urgency were factors associated with the occurrence of intraoperative complications. ASA III class, the presence of intraoperative complications, major surgery and duration of surgery greater than or equal to two hours were factors associated with intraoperative transfusion, whereas non-consumption of alcohol was protective against transfusion. The mean haemoglobin level was lower in the transfused group (10.4 g/dl) than in the non-transfused group (12.2 g/dl).

The mean age was 68.5 ± 6.9. The majority of patients were between 60 and 69 years of age, accounting for 60.6% of all patients. Our results are similar to those of many authors, in particular: Garcia-Casanovas [21] in a retrospective multicentre cohort study of 30,926 patients undergoing total knee or hip arthroplasty found a median age of 71 years. Ba EHB [22] in Dakar found a mean age of 74.9 years for elderly patients undergoing all types of surgery. Aubrun [1] found an average age of 67 ± 21 years, and the 2019 report by the French high health authority [23] found an average age of 65 ± 20 years. This reflects the reality of an ageing world population and their frequency in the operating theatre. The majority of patients were male, with a M/F sex ratio of 1.09. In contrast, Garcia-Casanovas [21] found that 57.1% of patients were female. The literature reports an over-representation of women, particularly in femoral neck fractures, in association with post-menopausal osteoporosis [24] . Perhaps, in our context, elderly women with hip fractures do not consult a doctor and stay at home with their fractures or treat them traditionally. Obesity affected 28% of patients, corroborating the findings of the WHO’s 2016 report, which showed that the prevalence of obesity is increasing worldwide.

It is likely that the western lifestyle adopted in our cities and the more westernised diet explain this fact [25] . Without any convincing explanation, there were no underweight patients in this series. In addition, cardiovascular (64.5%), renal and metabolic disorders (38.6%) and neuropsychiatric (9.4%) comorbidities predominated, in line with the literature on orthopaedic surgery in the elderly [1] [26] [27] . Aging is accompanied by an increased risk of comorbidity [8] . The elderly and their frequent co-morbidities have a high risk of VTE, and in our study, patients with a risk of VTE represented 83.3%. The presence of this thromboembolic risk was a factor associated with the occurrence of complications. The two deaths recorded were due to pulmonary embolism. This calls for a rigorous prophylactic strategy. In orthopaedic surgery, particularly in the elderly, guides lines recommend the use of low molecular weight heparins (LMWH) or direct oral anticoagulants, adapted to the patient’s profile and the type of operation [28] . In this series, despite systematic prevention with Enoxaparin (four hours after the end of surgery, 40 to 80 mg per day for at least three weeks, given the scarcity of direct oral anticoagulants on the local market), two fatal cases were recorded. In our study, ASA class II was predominant (69.7%). Our results are similar to those of Weickmans et al. [29] who found a predominance of ASA class 2 (81/100, 81%). The High health authority report in France [23] also noted a predominance (73%) of ASA class 2. This could be explained by the high prevalence of cardiovascular and metabolic co-morbidities which raise the ASA score. In our study, no biological variable showed a statistically significant difference between patients with or without intraoperative complications. These results suggest that in elderly patients undergoing orthopaedic surgery, intraoperative complications do not appear to be directly correlated with routine biological parameters. This corroborates the data in the literature which indicate that in this context, other factors such as preoperative functional status, co-morbid conditions (frailty, heart failure) or the complexity of the surgical procedure play a predominant role in the occurrence of intraoperative events [30] [31] . However, a significant difference was observed in haemoglobin levels in patients who received an intraoperative transfusion. Transfused patients had a significantly lower preoperative haemoglobin (10.4 ± 2.4 g/dl) than non-transfused patients (12.2 ± 1.8 g/dl; p < 0.001). This observation is in line with numerous studies, which identify preoperative anaemia as a major risk factor for transfusion, particularly in the context of major musculoskeletal surgery in the elderly [32] [33] . In this population, the proactive strategy of Patient Blood Management (PBM) is fully justified. For example, the correction of anaemia prior to orthopaedic surgery, particularly for femoral neck fractures or arthroplasty, is now strongly recommended [34] . Unfortunately, PBM has been slow to become established in the country because of local problems related to patient scheduling, collaboration with the surgeons and financial problems.

Locoregional anaesthesia was the technique used in 72% of cases, followed by GA in 23.4%. The anaesthetic agents most commonly used in LRA were bupivacaine and intrathecal morphine, sometimes combined with clonidine or fat-soluble opioids such as fentanyl. This combination aims to potentiate the analgesic effect while reducing the necessary doses of each agent, a strategy that is particularly beneficial in elderly patients, who are more sensitive to side effects [35] [36] . Fentanyl and sufentanyl were the morphinics used intraoperatively, and it should be noted that this use must remain cautious in the elderly patient, given the risk of respiratory depression and postoperative confusion [37] . It should be noted that short-acting morphines such as alfentanyl and remifentanyl are not available on the local market. The same applies to local anaesthetics such as levobupivacaine and ropivacaine. In addition, nearly 18% of patients required the administration of vasopressors, probably due to the vasodilatory effect of spinal anaesthesia, to which elderly patients are particularly sensitive due to their lower physiological reserve [38] . This frequency is consistent with the data in the literature on spinal anaesthesia in geriatric orthopaedics. The data in the literature indicate that one anaesthetic technique (GA vs ALR) should not be preferred in order to reduce mortality after surgery for fracture of the upper end of the femur), for example. There is currently no randomised prospective study to answer the question of a possible advantage in terms of mortality with general anaesthesia (GA) versus spinal anaesthesia (SA). There are contradictory retrospective studies sometimes finding an advantage in favour of SA and sometimes no difference [39] . The choice of anaesthetic technique depends on the habits of the intensive care anaesthetist, comorbidities and patient choice. It is probable that the lower cost and the ease of carrying out spinal anaesthesia militated in favour of this technique in a context of limited resources. The lack of experience and equipment has not permitted the practice of peripheral block. Intraoperative complications were observed in 31.8% of patients, a notable frequency in the context of orthopaedic surgery in elderly patients. Arterial hypotension was the most frequent complication (21.96%), probably related to spinal anaesthesia, the technique used in the majority of this series. This observation is in line with the data in the literature, in particular those reported by Capdevila [40] , who stresses the frequency of hypotension in elderly patients in association with a reduction in vascular tone and an altered haemodynamic response to locoregional anaesthesia. As hypotension is recognised as an important independent factor in the occurrence of complications and postoperative mortality [41] [42] , the French Society of Anaesthesia and Critical Care recommends that any intraoperative hypotension in elderly patients should be treated without delay in order to limit the risk of renal or myocardial complications [39] . This is why 18% of patients had benefited from vasopressors. Furthermore, the recourse to intraoperative transfusion in 19.7% of patients reflects the importance of blood loss in major orthopaedic surgery, but also the presence of anaemia preoperatively. This rate is comparable to that found in Carson’s work [43] , which shows a high prevalence of transfusion in elderly patients, often linked to a reduced physiological reserve and an increased prevalence of preoperative anaemia. Postoperative complications were limited (5.3%), but potentially serious, including two cases of fatal pulmonary embolism (1.5%). These results are in line with the observations of Anderson [44] , who reported that the risk of thromboembolism remains significantly increased after major orthopaedic surgery, despite preventive measures. Careful monitoring of these patients in the immediate postoperative period and well-managed anticoagulant prophylaxis therefore appear to be essential in limiting these serious events. However, despite the fact that prophylaxis with Enoxaparin was the rule in all patients with this risk, two cases of fatal pulmonary embolism were recorded. Only one case of agitation and no cases of delirium were encountered in this series, although this is a frequent complication in the elderly, contrary to the data in the literature [9] [10] . This may be explained by the fact that the elderly in our context live in families with children or grandchildren and not in old people’s homes as in high-income countries.

Our study highlights the preponderance of traumatic pathologies among the indications for surgery, with fractures of the hip (15.9%), femoral shaft (14.3%) and leg bones (12.8%) in the lead. These results corroborate data from the literature, which reports a high incidence of fractures of the lower limbs, particularly in elderly subjects, due to osteoporosis and accidental falls [45] . The frequency of hip fractures, in particular, is a public health issue in countries with limited resources, where specialised facilities are sometimes inadequate [46] . In terms of treatment, total hip replacement was the most common surgical procedure (25.7%), followed by osteosynthesis without external fixator (18.9%) and osteosynthesis with external fixator (12.1%). These results are in line with current recommendations, which favour arthroplasty for intracapsular femoral neck fractures in elderly patients to ensure optimal functional recovery [47] . Nevertheless, osteosynthesis remains the treatment of choice for diaphyseal fractures and certain open fractures, particularly in young patients or patients with polytrauma [48] . Chronic degenerative diseases such as coxarthrosis (12.1%) and gonarthrosis (11.3%) account for a large proportion of indications for surgery. The use of total hip or knee replacement in these cases reflects the technological advances and functional benefits demonstrated in several studies [49] . The presence of gangrene (5.3%) justifies recourse to mutilating procedures such as amputation or disarticulation (6%). This frequency highlights the importance of delays in treatment, which are common in low-resource countries where access to specialist care is limited [50] . Lastly, removal of osteosynthesis material accounts for 5.3% of operations, reflecting the importance of long-term post-operative follow-up. The literature emphasises that this type of procedure, although often secondary, remains essential in certain cases of chronic pain, infection or functional discomfort [51] . Several factors were found to be independently associated with the occurrence of intraoperative complications. Cardiovascular comorbidities (OR = 2.83; 95% CI: 1.12 - 7.15) are a major risk factor, in line with the data of Fleisher and Lee [52] [53] on the haemodynamic impact of major surgery in frail patients. The risk of VTE was also an important predictive factor (OR = 3.07; CI95%: 2.51 - 8.39), which underlines the need for systematic screening and appropriate prophylaxis in this population, particularly after hip or knee surgery, known for their high thromboembolic risk [54] . Furthermore, the performance of a major surgical procedure (OR = 2.61; IC95%: 1.70 - 4.80) and the urgency of the operation (OR = 4.89; IC95%: 2.68 - 6.90) are two factors strongly associated with an increased risk of complications. Urgent surgery, particularly for femoral neck fractures, is often associated with a lack of optimal preoperative preparation, which may explain the increased risk [55] . These results confirm the importance of a rigorous, individualised preoperative assessment in elderly patients, taking into account cardiovascular comorbidities, thromboembolic risk, the type of surgery envisaged and its emergency context. Anticipatory multidisciplinary management, including anaesthetists, geriatricians and surgeons, is essential to limit intraoperative complications in this high-risk population. The ASA III score was a risk factor for intraoperative transfusion, which is consistent with the data in the literature: in elderly patients, an altered general condition not only increases transfusion requirements but also limits tolerance of anaemia, due to reduced physiological reserves [56] [57] . Furthermore, the duration of surgery ≥ 2 hours, the occurrence of intraoperative complications and the major nature of the operation were also associated with an increased risk of transfusion. These factors reflect greater operative complexity, often correlated with increased bleeding, particularly in procedures such as arthroplasty or complex osteosynthesis [58] . These results are in line with those of Carson who found a correlation between operating time, blood loss and recourse to transfusion in orthopaedics in elderly patients [43] [55] . More originally, the absence of alcohol consumption was found to be a protective factor. Although this finding is less widely reported in the literature, it may be explained by the known effects of chronic alcoholism on liver function and coagulation. These alterations may predispose to haemorrhagic disorders, which are particularly deleterious in elderly patients who are often polypathological [59] . On the other hand, variables such as respiratory comorbidities or duration of anaesthesia were not associated with the risk of transfusion after adjustment. This suggests that these elements, although potentially correlated with an altered general state, are not independent factors but rather indirect markers of severity.

Overall, these results argue in favour of an individualised transfusion risk management strategy for orthopaedic surgery in the elderly. A Patient Blood Management approach integrating preoperative optimisation of haemoglobin, assessment of comorbidities and adapted planning of the surgical procedure seems essential to reduce the need for transfusions and their associated complications.

Strengths and Limitations of the Study

The strength of our study lies in the fact that it is the first to assess the factors associated with morbidity and mortality in orthopaedic surgery on elderly patients in our setting. This will help to improve the management of these patients and show practitioners that elderly patients currently benefit from many major orthopaedic procedures in our setting. Apart from its strengths, this study also has its limitations. The first limitation is the retrospective nature of the analysis. The information gathered is dependent on registry data. Some interesting information, such as the Lee score, was not found. The second limitation is linked to its mono-centric nature, which may limit the generalisability of the results.

The results of this study show that anaesthesia of elderly patients for orthopaedic procedures is becoming increasingly common, and that the risk of complications and even mortality is not zero. Good preoperative assessment of the haemodynamic status of these patients, together with rigorous and optimal management of the intraoperative and postoperative periods, will help to minimise this risk as much as possible. With the foreseeable increase in the number of elderly people in Africa in general and in the DRC in particular, it is important that their care is adapted to their needs, and it is probably necessary to advocate specialised multidisciplinary intraoperative management involving intensive care anaesthetists, surgeons, geriatricians, pharmacists and carers in order to improve the postoperative outcome of elderly patients undergoing orthopaedic surgery. A prospective study, therefore, seems useful, and an implementation study of patient blood management seems more than necessary.

Patrick Kintieti: Study design, data collection and drafting of the manuscript; Wilffrid Mbombo: Study design, data collection and drafting of the manuscript; Alphonse Mosolo: data collection and correction of the manuscript; Aliocha Nkodila: Statistical analysis; All other authors: correction of the manuscript.

We would like to thank the entire operating theatre team at Monkole Hospital for their collaboration: Bénoit Rwabahizi, Cathy Nziavake, Jacquies Byengangu, Maguy Tshiabu, Olga Milo, Samuel Makengo and Sylvie Ngunda.