Comparative Efficacy of Neuraxial and General anesthesia for hip fracture surgery: A meta-analysis of randomized clinical trials

DOI: https://doi.org/10.21203/rs.2.17433/v3

Abstract

Background: The choice of anesthesia technique remains debatable in patients undergoing surgical repair of hip fracture. This meta-analysis was performed to compare the effect of neuraxial (epidural/spinal) versus general anesthesia on perioperative outcomes in patients undergoing hip fracture surgery.

Methods: Medline, Cochrane Library, Science-Direct, and EMBASE databases were searched to identify eligible studies focused on the comparison between neuraxial and general anesthesia in hip fracture patients between January 2000 and May 2019. Perioperative outcomes were extracted for systemic analysis. The sensitivity analyses were performed by the leave-one-out approach. The evidence quality for each outcome was evaluated by the Grading of Recommendations Assessment, Development and Evaluation (GRADE) system.

Results: Nine randomized controlled trials (RCTs) including 1084 patients fulfilled our selection criteria. The results showed that there were no significant differences in the 30-day mortality (OR = 1.34, 95% CI 0.56, 3.21; P = 0.51), length of stay (MD = -0.65, 95% CI -0.32, 0.02; P =0.06), and the prevalence of delirium (OR = 1.05, 95% CI 0.27, 4.00; P = 0.95), acute myocardial infarction (OR = 0.88, 95% CI 0.17, 4.65; P = 0.88), deep venous thrombosis (OR = 0.48, 95% CI 0.09, 2.72; P = 0.41), and pneumonia (OR = 1.04, 95% CI 0.23, 4.61; P = 0.96) for neuraxial anesthesia compared to general anesthesia. There was a significant difference in terms of blood loss in favor of the neuraxial anesthesia (MD = -137.8, 95% CI -241.49, -34.12; p = 0.009). The evidence quality for each outcome evaluated by the GRADE system was low.

Conclusions: In summary, our present study demonstrated that neuraxial anesthesia is associated with a reduced blood loss in patients undergoing hip fracture surgery compared to general anesthesia. However, this result was underpowered. Due to small sample size and enormous inconsistency in the choice of outcome measures, more high-quality studies with large sample size are needed to clarify this issue

Background

Hip fracture is one of the most common injuries that occurs in about 1.6 million people around the world each year; the number is estimated to reach more than six million by 20501. Moreover, there are a range of comorbidities in elderly patients with hip fracture, which are associated with an increased risk of morbidity and mortality2. Most hip fractures should be treated surgically that requires some type of anesthesia3.

Thus far, the ideal choice between neuraxial and general anesthesia has not been identified. Several studies demonstrated that compared with general anesthesia, neuraxial anesthesia has some advantages such as airway management avoidance, no intubation requirement, and prolonged postoperative analgesia4.  Furthermore, neuraxial anesthesia could provide excellent intraoperative anesthesia, decrease blood loss, potentially reduce risk of postoperative nausea and vomitting (PONV), as well as deep venous thrombosis4-6. Conversely, general anesthesia is reported to provide a more stable hemodynamic state, faster induction, and avoid some complications such as pneumonia, epidural haematoma and infection7,8. However, the effect of the two anesthesia techniques on patients with hip fracture is controversial regarding postoperative outcomes. A recent systematic review including 15 studies revealed that neuraxial anesthesia was only associated with a shorter length of hospital stay in patients undergoing hip fracture surgery. This review emphasized that sensitivity analyses showed marginal statistical significance for length of stay favoring spinal anaesthesia, and the definitions of reported outcomes varied widely or were unclear, making evaluation in a standardized manner very difficult.9 Another systematic review reported a reduced in-hospital mortality in the neuraxial anaesthesia group, but no definitive conclusion can be drawn for longer-term mortality10. Both of them have recommended that further high-quality studies be performed.

To date, several most recent randomized controlled trials (RCTs) have been published, which assessed the effect of the two anesthesia techniques for hip fracture surgery. Through including these RCTs, our study aimed to systematically evaluate perioperative outcomes of patients with hip fracture surgery, and provide more reliable evidence to identify the optimal technique.

Methods

This meta-analysis was conducted in accordance with Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines. It was registered in the international prospective register of systematic reviews (Prospero: CRD42020143172).

Search Strategy

Medline, Cochrane Library, Science-Direct, and EMBASE databases were searched by two independent reviewers between January 2000 and May 2019. We selected studies of neuraxial anesthesia compared with general anesthesia in patients undergoing hip fracture surgery. Following iterms were searched for both alone and in various combinations, “hip fracture” or “femur fracture” or “intertrochanteric” or “femoral neck” AND “regional anesthesia” or “spinal anesthesia” or “neuraxial anesthesia” or “epidural anesthesia”. The “related articles” function in Medline was performed to expand the search. Reference lists were also hand-searched for relevant studies. No language restriction was placed on our search. 

Inclusion and Exclusion Criteria

Two independent reviewers screened article titles and abstracts based on the following inclusion criteria: (1) randomized controlled trials (RCTs) with no language restriction; (2) studies comparing general anesthesia with neuraxial anesthesia (epidural or spinal) in patients undergoing hip fracture surgery; (3) studies provided numerical data. The following exclusion criteria was used: (1) studies that did not meet the inclusion criteria; (2) unpublished data or repeated data; (3) abstracts, case reports, comments, conference papers, or animal studies, meta-analysis and systematic reviews. 

Data Extraction

Two independent reviewers designed a structured table and collected all the relevant data into a database. The following information was extracted from each study that met the inclusion criteria: first author’s name, publication year, country, sample size, age, American Society of Anesthesiologists (ASA) physical status, anesthesia technique, surgery type, study outcome measures. We also attempted to contact the corresponding authors to verify the accuracy of the data and to obtain further analytical data. We performed a meta-analysis for blood loss, 30-day mortality, length of hospital stay, and the prevalence of delirium, acute myocardial infarction, deep venous thrombosis, and pneumonia. 

Methodological Quality Assessment

The methodological quality of each RCT was assessed using the Cochrane Handbook for Systematic Reviews of Interventions 5.1 by two reviewers, which contained the following items: random sequence generation, allocation concealment, blinding, incomplete outcome data, selective reporting, and other sources of bias. It was judged by answering a question, with “yes” indicating low risk of bias, “no” indicating high risk of bias, and “unclear” indicating unclear or unknown risk of bias11 . The corresponding author was also consulted when any disagreement exists, and a consensus was reached by discussion. 

Statistical Analysis

The statistical analysis of the pooled data were performed using Review Manager software (version 5.1, The Cochrane Collaboration, Oxford, England). For continuous variables, standardized mean difference (SMD) or weighted mean (WMD) difference was calculated with the 95% confidence intervals (CIs) as a summary statistic. For dichotomous variables, relative risk (RR) and 95% CIs were used. The combined effect was considered significant at a 2-sided P < 0.05. The p-value with the Cochrane Q-test was texted, and the I2 statistic was used to judge inconsistency of treatment effects across studies. A random effect model was used if high heterogeneity was detected (p<0.10, I2>50%); otherwise, a fixed effect model was used if low heterogeneity existed (p>0.10, I2<50%). Sensitivity analyses were performed by leave-one-out approach by STATA 14.0 (STATA Corp, College Station, TX). Bonferroni correction was applied for multiple comparisons. Publication bias was evaluated by funnel plot, if our meta-analysis included more than 10 studies12

Evidence synthesis

The evidence grade for the main outcomes are assessed using the guidelines of the (GRADE) system working group including the following items: risk of bias, inconsistency, indirectness, imprecision and publication bias. The recommendation level of evidence is classified into the following categories: (1) high, which means that further research is unlikely to change confidence in the effect estimate; (2) moderate, which means that further research is likely to significantly change confidence in the effect estimate but may change the estimate; (3) low, which means that further research is likely to significantly change confidence in the effect estimate and to change the estimate; and (4) very low, which means that any effect estimate is uncertain. The evidence quality is graded using the GRADEpro Version 3.6 software. The evidence quality was graded using the GRADEpro Version 3.6 software. The strengths of the recommendations were based on the quality of the evidence.

Results

Study identification and selection

A total of 1274 relevant studies were identified according to the search strategy. However, 798 publications were excluded after checking for duplicates. Among the 476 remaining articles, 359 articles were excluded after reviewing the titles and abstracts. Then we assessed 17 studies with full texts for eligibility. Eight studies were excluded because four of them included no control groups, and others provided inadequate data. Finally, nine RCTs with a total of 1084 patients between 2003 and 2018 met our inclusion criteria, and were included in the meta-analysis13-21. The flow diagram of study selection is shown in Fig. 1.

 

Study Characteristics

All the included studies were written in English, which examined perioperative outcomes between hip fracture patients who receive neuraxial or general anesthesia undergoing surgical repair. There were a total of 1084 patients, whose ages were older than 49 years-old. Seven studies looked at outcomes relating to spinal anesthesia compared with general anesthesia13,14,16-19,21, one study examed outcomes for hypobaric unilateral spinal anesthesia and general anesthesia20, and the other study compared general versus neuraxial anesthesia that encompassed spinal and epidural anesthesia15. In the terms of surgery type, two studies performed arthroplasty, hip screw and intramedullary nail18,20; two studies included hemiarthroplasty only13,14, and one study performed hemiarthroplasty and Intramedullary nail16. Only one study was at a high risk of performance bias14, and the other studies were all at low risk or unclear (Fig. 2). The characteristics of the included studies is shown in Table 1.

 

Outcomes for meta-analysis

Delirium rate was reported in four studies with 400 patients in the neuraxial anesthesia group and 409 patients in the general anesthesia group13,15,18,21. The P value with the Cochran’s Q test was 0.03, and the I2 statistic was 66%, which indicated high heterogeneity among these studies. Thus a random effect model was used to analyze the results. The pooled data showed no significant difference in delirium rate between the two groups (OR = 1.05, 95% CI 0.27, 4.00; P = 0.95, Fig. 3)

Three studies examed blood loss during hip fracture surgery with 250 patients in the neuraxial anesthesia group and 257 patients in the general anesthesia group15,17,19. The P value with the Cochran’s Q test was 0.0003, and the I2 statistic was 88%, which indicated high heterogeneity among these studies. Thus a random effect model was used to analyze the results, showing a significant difference between the two groups in favor of the neuraxial anesthesia (MD = -137.8, 95% CI -241.49, -34.12; p = 0.009, Fig. 4). However, the adjusted p-value was 0.054 with a Bonferroni correction applied and therefore above the significant threshold of 0.05. Thus, this result was underpowered.

Three studies were included in the meta-analysis for 30-day mortality, involving 363 patients in the neuraxial anesthesia group and 389 patients in the general anesthesia group15,16,18. The P value with the Cochran’s Q test was 0.21, and the I2 statistic was 48%, which indicated low heterogeneity among these studies.Thus a fixed effect model was used to analyze the results. The pooled data revealed that there was no significant difference in 30-day mortality between the two groups (OR = 1.34, 95% CI 0.56, 3.21; P = 0.51, Fig. 5)

Acute myocardial infarction rate was reported in three studies with 363 patients in the neuraxial anesthesia group and 376 patients in the general anesthesia group15,16,18. The P value with the Cochran’s Q test was 0.96, and the I2 statistic was 0%, which indicated low heterogeneity among these studies. Thus a fixed effect model was used to analyze the results. The pooled data showed no significant difference in the acute myocardial infarction rate between the two groups (OR = 0.88, 95% CI 0.17, 4.65; P = 0.88, Fig. 6)

Two studies provided the outcome of pneumonia rate, which involved 363 patients in the neuraxial anesthesia group and 389 patients in the general anesthesia group15,18. The P value with the Cochran’s Q test was 0.42, and the I2 statistic was 0%, which indicated low heterogeneity among these studies. Thus a fixed effect model was used to analyze the results. The pooled data showed no significant difference in pneumonia rate between the two groups (OR = 1.04, 95% CI 0.23, 4.61; P = 0.96, Fig. 7)

Two studies reported length of stay in a way that could be comparable by meta-analysis, including 348 patients in the neuraxial anesthesia group and 361 patients in the general anesthesia group15,18. The P value with the Cochran’s Q test was 0.54, and the I2 statistic was 0%, which indicated low heterogeneity among these studies. Thus a fixed effect model was used to analyze the results.The pooled data revealed that no significant difference was detected in the length of stay between the two groups (MD = -0.65, 95% CI -0.32, 0.02; P =0.06, Fig. 8)

Two studies were included in the meta-analysis for deep venous thrombosis rate, involving 179 patients in the neuraxial anesthesia group and 183 patients in the general anesthesia group18,20. The P value with the Cochran’s Q test was 0.60, and the I2 statistic was 0%, which indicated low heterogeneity among these studies. Thus a fixed effect model was used to analyze the results. The pooled data revealed that there was no significant difference in deep venous thrombosis rate between the two groups (OR = 0.48, 95% CI 0.09, 2.72; P = 0.41, Fig. 9)

Sensitivity analysis

Sensitivity analyses were performed by the leave-one-out approach from the aforementioned meta-analyses. No difference was detected in the direction of the conclusions with each study removed in turn, which showed that our study had good reliability, and the results were statistically robust (Fig. 9)..

Quality of the evidence and recommendation strengths

A total of seven outcomes in this meta-analysis were evaluated using the GRADE system (Table 2). The evidence quality for each outcome was low. Therefore, we demonstrate that the overall evidence quality is low, which means that further research is likely to significantly change confidence in the effect estimate and may change the estimate.

Discussion

In our study, a total of nine RCTs with 1084 patients were included to make an updated meta-analysis. However, no significant difference was detected in the 30-day mortality, length of stay, and the prevalence of delirium, acute myocardial infarction, and pneumonia in patients undergoing hip fracture surgery where either neuraxial or general anesthesia was used. We first focused on the comparison of blood loss between the two anesthesia techniques. We found that neuraxial anesthesia was associated with a reduced blood loss. The sensitivity analysis showed that the result had good reliability. Nevertheless, this result was assessed underpowered by Bonferroni correction. The sample size was also small, and the overall evidence was low, indicating that further research is likely to significantly change confidence in the effect estimate and may change the estimate. Based on the current available evidence, more high-quality RCTs are required for further investigation.

According to methodological quality assessment, eight out of nine RCTs in our study were assessed as high-quality. Moreover, our study included several RCTs, in which the results were published after the most recent systematic review of this topic, making our results more dependable. Of note, all of the RCTs showed low risk of attrition bias and reporting bias that may contribute to reducing systematic bias. Another strength of our study is low heterogeneity, detected in five out of six outcome measures assessed using I2 statistic, demonstrating consistent outcomes across the comparisons. In addition, the adjustment was made by Bonferroni correction to decrease the risk of type I error caused by multiple statistical tests in our study. Also, some data of previous reviews dated back to the 1980s9,10, in which the type of anaesthetic techniques may not reflect current clinical practice, and it may restrain us from finding clinically relevant differences between the two techniques9,46, while our study included most recent RCTs.

According to pharmacology, neuraxial anesthesia could lead to lower heart rate, and blood pressure than general anesthesia by blocking alpha and beta adrenergic receptors. Consequently, controlled blood pressure resulted in intraoperative less blood loss in neuraxial anesthesia patients16,28. Current practice revealed that the number of patients who needed blood transfusion was larger in general anesthesia group, which means patients receiving spinal anesthesia had less blood loss than those receiving general anesthesia19,29,30. In consistency with this result, a systematic review by Richman et al. Including 66 articles demonstrated that the use of neuraxial anesthesia resulted in a significant decrease in estimated blood loss31. However, a meta-analysis by Hu er al. including 21 RCTs stated that there was insufficient evidence to support the use of neuraxial anesthesia in decreasing intraoperative blood loss32. In our study, only three RCTs involving 507 patients have been summarised. Two of them showed the neuraxial anesthesia was assosiated with statistically significant decrease in blood loss, the other showed no significant difference between the two anesthesia techniques. However, the results from our meta-analysis indicating decreased blood loss with neuraxial anesthesia are limited by a high degree of heterogeneity (88%) and low-quality evidence for this outcome. Also, we did not investigate whether this resulted in a clinically meaningful difference in perioperative blood transfusions.

Delirium is a very common postoperative complication, which leads to lasting cognitive and functional decline, and increasing length of stay18,22. There are many precipitating factors in developing delirium, including infection, myocardial and cerebral ischaemia, urinary retention, pain, constipation as well as electrolyte abnormalities23. Furthermore, several studies have investigated the incidence of delirium in elderly patients, who were admitted to be hospitalized for a variety of reasons, and the prevalence amongst medical wards was estimated to range from 29% to 64%23-25. Additionally, the development of delirium is thought to be multifactorial process. Certain patient characteristics are also easy to cause delirium, including pre-existing cognitive impairment, sleep deprivation, medical immobilities, visual impairment, hearing impairment and poly pharmacy26,27. Our study detected no significant difference in delirium rate between general and neuraxial anesthesia. It is noteworthy that none of the included studies represented relative characteristics and potential risk factors that causing delirium in hip fracture patients perioperatively. Thus the result may be unconvincing.

Our study detect comparable outcomes in the incidence of 30-day mortality between the two groups. In line with this result, a retrospective study reported that the anesthesia technique has little effect on postoperative mortality, and the type of anesthesia given by the anesthesiologist should be selected based on the individual physical condition33. The study of Lienhart et al. including 425 patients indicated that their coexisting disease has great influence on 30-day mortality in old patients such as diabetes, cardiovascular disease, etc34. Delay of surgery for more than 24h was a main factor affecting postoperative mortality in geriatric hip fracture patients35. The retrospective cohort study of Pincus et al. Investigated 42230 patients undergoing hip fracture surgery, and demonstrate that a preoperative waiting time of more than 24h was associated with a greater risk of 30-day mortality and other complications36.

In our study, the incidence of myocardial infarction and pneumonia were similar in both groups. Zuo et al. detected the same result, and suggested that the neuraxial anesthesia might be a better choice in hip fracture surgery37. However, Urwin et al. proposed that the incidence of myocardial infarction and pneumonia was lower in patients receiving neuraxial anesthesia, and a significant lower incidence of intraoperative hypotension was detected in patients receiving general anesthesia38. It should be noted that Urwin et al. evaluated 2161 patients retrospectively. Moreover, all of the included studies were performed more than 20 years ago, which are now somewhat dated, since many drugs used for anesthesia techniques and health care systems have been improved a lot. Thus their findings could not provide worthy references to some extent.

There was no significant difference regarding the length of stay between the two anesthesia techniques. Sutcliffe et al. surveyed 1333 volunteers of hip surgery, and found no difference in factors of hospitalization in both groups39. Neverthless, Neuman et al. conducted a matched retrospective cohort study involving 56729 patients, and found a modestly shorter length of stay in the neuraxial anesthesia group. The authors also proposed that the fracture type and performed surgery procedure were important factors; minimally invasive approaches and optimal quality of fracture reduction may decrease the length of stay40. In addition, Grant et al. declared that the pain severity was lower in patients receiving general anesthesia, resuting in shorter length of stay41. A notable point is that waiting time prior to surgery extended the length of stay42. In our meta-analysis, one study reported the overall length of stay18 while the other documented the length of stay before and after the surgery15. It is difficult to draw a definite conclusion due to the existence of aforementioned multiple factors. Also, the small sample size in our study should be taken into consideration.

Perioperative deep venous thrombosis is common in hip fracture patients. Several studies concluded that neuraxial anaesthesia was associated with fewer incidents of deep venous thrombosis when compared to general anaesthesia38,43,44. It was thought that in neuraxial anaesthesia sympathetic block could lead to vasodilatation of the lower limbs, and then the increased blood flow to the lower limbs was likely to reduce the coagulability and viscosity of blood45. A Cochrane review published in 2016 by pooling the results from 31 RCTs showed a reduced risk of deep venous thrombosis in the neuraxial group without potent thromboprophylaxis. Nevertheless, the level of evidence was very low for this outcome46. Another Cochrane review concluded that there was a marginal advantage for neuraxial anaesthesia regarding the incidence of deep venous thrombosis47. Our study included only two RCTs, and detected no significant difference in the incidence of deep venous thrombosis between the two groups.

Objectively speaking, several limitations of our study should be mentioned. A major limitation is that the sample size was relatively small, and the sample size varied widely among the included studies. Another notable limitation is that most of the included studies did not describe whether additional sedation was used in hip fracture patients receiving neuraxial anesthesia, for instance, the use of propofol sedation could influence the prevalence of postoperative delirium48. Also, no information is available in the terms of the dosage of the anesthesia used. In addition, the inconsistent definition of length of stay and delirium may account for the wide prevalence range for these outcomes. There are numerous confounding factors such as the diversity of patient groups, health care systems, surgical and anesthetic techniques that may affect the perioperative outcomes, leading to potential biases. This issue would be possibly considered as a weakness. Last but not least, the outcome measures were not identical in each trial, thus we did not have sufficient data to perform other meta-analyses, which potentially affects the current findings of our study. Therefore, more high-quality RCTs with large sample size are required for a firm conclusion.

Conclusion

In summary, our present study demonstrated that neuraxial anesthesia is associated with a reduced blood loss in patients undergoing hip fracture surgery when compared to general anesthesia, however, it was assessed as a low-quality evidence. We suggest that the choice of anaesthesia (neuraxial or general) should be made by the anaesthesiologist based on the individual patient’s requirements, comorbidities, potential postoperative complications, consultation of geriatrician and orthopaedic surgeon, and the clinical experience of the anaesthesiologist. Due to small sample size and enormous inconsistency in the choice of outcome measures, more high-quality studies with large sample size are needed to to clarify this issue.

Abbreviations

ASA: American Society of Anesthesiologists; CI: Confidence interval; GRADE: The grading of recommendations, assessment, development, and evaluation methodology; GA: general anesthesia; HUSA: hypobaric unilateral spinal anesthesia; MMSE: Mini Mental State Examination; NA: neuraxial anesthesia; PONV: Postoperative nausea and vomiting; PRISMA: Preferred reporting items for systematic reviews and meta-analyses; RCT: Randomized controlled trial; RR: Risk ratio; SMD: Standardized mean difference; WMD: weighted mean difference; DVT, deep venous thrombosis

Declarations

Ethics approval and consent to participate

Not applicable. 

Consent for publication

Not applicable. 

Availability of data and materials

All data generated or analyzed during this study are included in this published article. 

Competing interests

The authors declare that they have no competing interests. 

Funding

There was no funding source in this study. 

Authors’ contributions

XZ and ZL designed and conceived the study, performed the statistical analysis, and drafted the manuscript. YG participated in the interpretation of data, analysis, and drafting of the manuscript. YT participated in the study design and coordination, and helped to draft the manuscript. All authors read and approved the final manuscript.

Acknowledgments

Not applicable.

References

  1. Dhanwal DK, Dennison EM, Harvey NC, Cooper C. Epidemiology of hip fracture: worldwide geographic variation. Indian J Orthop. 2011;45:15-22.
  2. Hickson LJ, Farah WH, Johnson RL, et al. Death and Postoperative Complications After Hip Fracture Repair: Dialysis Effect.Kidney Int Rep. 2018;3:1294-1303.
  3. White SM. anesthesia type and cement reactions in hip fracture surgery. anesthesia. 2019;74:543-544.
  4. White SM,Foss NB.anesthesia aspects in the treatment of fragility fracture patients. Injury. 2018;49:1403-1408.
  5. Radić N, Radinović K, Ille M, et al. The selection of best anesthesiological technique for hip fracture surgery in older high-risk patients. Acta Chir Iugosl. 2012;59:113-115.
  6. Kowark A, Rossaint R. General versus spinal anesthesia for the elderly hip fractured patient. Curr Opin Anaesthesiol. 2019;32:116-119.
  7. Rivas E. Geriatric patients undergoing non-elective surgery for hip fracture: Can management be optimized? J Clin Anesth. 2019;59:112-113.
  8. Zhong H, Wang Y, Wang Y. Comparison of the effect and clinical value in general anesthesia and combined spinal-epidural anesthesia in elderly patients undergoing hip arthroplasty. Exp Ther Med. 2019;17:4421-4426.
  9. O'Donnell CM, McLoughlin L, Patterson CC, et al. Perioperative outcomes in the context of mode of anesthesia for patients undergoing hip fracture surgery: systematic review and meta-analysis. Br J Anaesth. 2018;120:37-50.
  10. Waesberghe JV, Stevanovic A, Rossaint R, et al. General vs. neuraxial anaesthesia in hip fracture patients: a systematic review and meta-analysis. BMC Anesthesiol. 2017;17(1):87.
  11. Higgins J, Green SE. Cochranehandbookfor systematic reviews of interventions version 5.1.0. (updated March2011). Available from: http://handbook-5-1.cochrane.org/
  12. Sterne JA, Higgins JP. Recommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised controlled trials. Bmj. 2011; 343(5):d4002
  13. Casati A, Aldegheri G, Vinciguerra E, et al. Randomized comparison between sevoflurane anesthesia and unilateral spinal anesthesia in elderly patients undergoing orthopaedic surgery. Eur J Anaesthesiol. 2003;20(8):640-646.
  14. Hoppenstein D, Zohar E, Ramaty E, et al. The effects of general vs spinal anesthesia on frontal cerebral oxygen saturation in geriatric patients undergoing emergency surgical fixation of the neck of femur. J Clin Anesth. 2005;17(6):0-438.
  15. Heidari SM, Soltani H, Hashemi SJ, et al. Comparative study of two anesthesia methods according to postoperative complications and one month mortality rate in the candidates of hip surgery. J Res Med Sci. 2011;16(3):323-330.
  16. Biboulet P, Jourdan A, Van Haevre V, et al. Hemodynamic Profile of Target-Controlled Spinal Anesthesia Compared With 2 Target-Controlled General Anesthesia Techniques in Elderly Patients With Cardiac Comorbidities. Reg Anesth Pain Med. 2012;37(4):433-440.
  17. Messina A, Frassanito L, Colombo D, et al. Hemodynamic changes associated with spinal and general anesthesia for hip fracture surgery in severe ASA III elderly population: a pilot trial. Minerva Anestesiol. 2013;79(9):1021-1029.
  18. Parker MJ, Griffiths R. General versus regional anesthesia for hip fractures. A pilot randomised controlled trial of 322 patients. Injury.2015;46(8):1562-1566.
  19. Haghighi M, Sedighinejad A, Nabi BN, et al. Is spinal anesthesia with low dose lidocaine better than sevoflorane anesthesia in patients undergoing hip fracture surgery. Arch Bone Jt Surg. 2017; 5(4):226.
  20. Meuret P, Bouvet L, Villet B, et al. Hypobaric Unilateral Spinal anesthesia versus General anesthesia in Elderly Patients Undergoing Hip Fracture Surgical Repair: A Prospective Randomised Open Trial. Turk J Anaesthesiol Reanim. 2018;46(2):121-130.
  21. Tzimas P, Samara E, Petrou A, et al. The influence of anesthetic techniques on postoperative cognitive function in elderly patients undergoing hip fracture surgery: General vs Spinal Anesthesia. Injury. 2018;49(12):2221-2226.
  22. Schubert M, Schürch R, Boettger S, et al. A hospital-wide evaluation of delirium prevalence and outcomes in acute care patients - a cohort study. Bmc Health Services Research. 2018; 18(1):550.
  23. Elie M, Cole M G, Primeau F J, et al. Delirium Risk Factors in Elderly Hospitalized Patients. J Gen Intern Med. 2010;13(3):204-212.
  24. Holt R, Young J, Heseltine D. Effectiveness of a multi-component intervention to reduce delirium incidence in elderly care wards. Age Ageing. 2013;42(6):721-7.
  25. A. Hernandez, Lindroth H, Rowley P, et al. Post-anesthesia care unit delirium: incidence, risk factors and associated adverse outcomes. Br J Anaesth. 2017;119(2):288-290.
  26. Smith T O, Cooper A, Peryer G, et al. Factors predicting incidence of post-operative delirium in older people following hip fracture surgery: a systematic review and meta-analysis. Int J Geriatr Psychiatry. 2017;32(4):386-396.
  27. Pessoa RF, Nácul FE. Delirium in the critically ill patient. Rev Bras Ter Intensiva. 2006; 18(2):190.
  28. Ing C, Sun L S, Friend AF, et al. Differences in intraoperative hemodynamics between spinal and general anesthesia in infants undergoing pyloromyotomy. Paediatr Anaesth. 2017;27(7).
  29. Liu JL, Xiao LW, Mao WG, et al. Comparative outcomes of peripheral nerve blocks versus general anesthesia for hip fractures in geriatric Chinese patients. Patient Prefer Adherence. 2014;8:651-9.
  30. Rosberg B, Fredin H, Gustafson C. Anesthetic techniques and surgical blood loss in total hip arthroplasty.[J]. Acta Anaesthesiol Scand. 2010;26(3):189-193.
  31. Richman JM, Rowlingson AJ, Maine DN, et al. Does neuraxial anesthesia reduce intraoperative blood loss?: A meta-analysis. J Clin Anesth. 2006; 18(6):427-35.
  32. Hu S, Zhang ZY, Hua YQ, et al. A comparison of regional and general anesthesia for total replacement of the hip or knee: A meta-analysis. J Bone Joint Surg Br. 2009;91(7):935-42.
  33. Chu CL, Chiou HY, Chou WH, et al. Leading Comorbidity associated with 30-day post-anesthetic mortality in geriatric surgical patients in Taiwan: a retrospective study from the health insurance data. Bmc Geriatrics. 2017;17(1):245.
  34. Lienhart A, Auroy Y, Péquignot F, et al. Survey of anesthesia-related mortality in France. Anesthesiology. 2006;105(6):1087-1097.
  35. Carretta E, Bochicchio V, Rucci P, et al. Hip fracture: effectiveness of early surgery to prevent 30-day mortality. Int Orthop. 2011;35(3):419.
  36. Pincus D, Ravi B, Wasserstein D, et al. Association Between Wait Time and 30-Day Mortality in Adults Undergoing Hip Fracture Surgery. JAMA. 2017;318(20):1994.
  37. Zuo D, Jin C, Shan M, et al. A comparison of general versus regional anesthesia for hip fracture surgery: a meta-analysis. Int J Clin Exp Med. 2015;8(11):20295-20301.
  38. Urwin SC, Parker MJ, Griffiths R. General versus regional anesthesia for hip fracture surgery: a meta- analysis of randomized trials. Br J Anaesth. 2000;84(4):450-455.
  39. Sutcliffe AJ, Parker M. Mortality after spinal and general anesthesia for surgical fixation of hip fractures. anesthesia. 1994;49(3):237-240.
  40. Neuman MD, Rosenbaum PR, Ludwig JM, et al. Anesthesia Technique, Mortality, and Length of Stay After Hip Fracture Surgery. JAMA. 2014;311(24):2508-2517.
  41. Grant MC, Pio Roda CM, Canner JK, et al. The Impact of Anesthesia-Influenced Process Measure Compliance on Length of Stay: Results From an Enhanced Recovery After Surgery for Colorectal Surgery Cohort. Anesth Analg. 2019;128(1):68-74.
  42. Shetty V, Vowlar S, Villar R. The Role of Anesthesia and the Anesthetist in Reducing the Length of Stay After Primary Total Hip Arthroplasty. J Arthroplasty. 2007;22(2):305-305.
  43. Macario A, Dexter F. What are the most important risk factors for a patient’s developing intraoperative hypothermia? Anesth Analg. 2002;94(1):215-220.
  44. Breen P, Park KW. General anesthesia versus regional anesthesia. Int Anesthesiol Clin. 2002;40(1):61-71.
  45. Davis FM, Quince M, Laurenson VG. Deep vein thrombosis and anaesthetic technique in emergency hip surgery. BMJ. 1980;281(6254):1528-1529.
  46. Guay J, Parker MJ, Gajendragadkar PR, Kopp S. Anaesthesia for hip fracture surgery in adults. Cochrane Database Syst Rev. 2016;2:CD000521.
  47. Parker MJ, Handoll HHG, Griffiths R. Anaesthesia for hip fracture surgery in adults. Cochrane Database Syst Rev. 2004;4:CD000521.
  48. Sieber FE, Zakriya KJ, Gottschalk A, et al.Sedation depth during spinal anesthesia and the development of postoperative delirium in elderly patients undergoing hip fracture repair. Mayo Clin Proc. 2010;85:18-26.

Tables

Table 1 The descriptive characteristics of included studies.

Study

Country

Sample size(male/female)

Age

in years

ASA status

anesthesia

surgery type

Study outcome measures

Casati et al. (2003)13

Italy

30 (7/23)

 

84

Ⅱ-Ⅲ

GA vs Spinal

Hemiarthroplasty

Delirium; MMSE

Hoppensteinet al. (2005)14

Israel

60

>65

-Ⅲ

GA vs Spinal

Hemiarthroplasty

Hemodynamic change; Delirium

 

Heidariet al. (2011)15

Iran

387 (257/130)

>60

-Ⅲ

GA vs NA

--

Length of stay;

30-day mortality;

Acute myocardial infarction; Pneumonia; Blood loss

 

 

Biboulet et al. (2012)16

France

45 (14/31)

>75

Ⅲ-Ⅳ

GA vs Spinal

Hemiarthroplasty

Intramedullary nail

30-day mortality;

Acute myocardial infarction

 

Messina et al. (2013)17

Italy

20 (7/13)

>75

GA vs Spinal

--

Blood loss; Hemodynamic change

Parker et al. (2015)18

UK

322 (87/235)

>49

-Ⅲ

GA vs Spinal

Arthroplasty

Sliding hip screw

Intramedullary nail

Delirium; 30-day mortality

Acute myocardial Infarction; Pneumonia; Length of stay;

DVP

 

 

 

 

Haghighi et al. (2017)19

Iran

100 (80/20)

>60

-Ⅲ

GA vs Spinal

--

Blood loss; PONV

Meuret et al. (2018)20

France

40 (8/32)

>75

-Ⅲ

GA vs HUSA

Arthroplasty

Dynamic hip screw

Intramedullary nail

PONV; DVP

Tzimas et al. (2018)21

Greece

 

70 (33/37)

76

-Ⅲ

GA vs Spinal

--

Delirium; MMSE

 

MMSE, Mini Mental State Examination; PONV, Post Operative Nausea And Vomitting; GA, general anesthesia; NA, neuraxial anesthesia; ASA, American Society of Anesthesiologists; HUSA, hypobaric unilateral spinal anesthesia; DVP, deep venous thrombosis.

 

Table 2 The GRADE evidence quality for main outcome.

Quality assessment

No of patients

Effect

Quality

Importance

 
 

No of studies

Design

Risk of bias

Inconsistency

Indirectness

Imprecision

Other considerations

Delirium

Control

Relative
(95% CI)

Absolute

 

Delirium rate

 

4

randomised trials

serious

no serious inconsistency

serious

no serious imprecision

none

26/400
(6.5%)

33/409
(8.1%)

OR 1.05 (0.27 to 4)

4 more per 1000 (from 58 fewer to 179 more)

ÅÅOO
LOW

IMPORTANT

 

 

12.7%

6 more per 1000 (from 89 fewer to 241 more)

 

Acute myocardial infarction rate

 

3

randomised trials

no serious risk of bias

serious

serious

no serious imprecision

none

2/363
(0.55%)

3/376
(0.8%)

OR 0.73 (0.14 to 3.74)

2 fewer per 1000 (from 7 fewer to 21 more)

ÅÅOO
LOW

IMPORTANT

 

 

0.6%

2 fewer per 1000 (from 5 fewer to 16 more)

 

30-day Mortality

 

3

randomised trials

serious

no serious inconsistency

no serious indirectness

serious

none

11/331
(3.3%)

16/343
(4.7%)

OR 0.71 (0.33 to 1.53)

13 fewer per 1000 (from 31 fewer to 23 more)

ÅÅOO
LOW

IMPORTANT

 

 

4.9%

14 fewer per 1000 (from 32 fewer to 24 more)

 

Blood loss (Better indicated by lower values)

 

3

randomised trials

serious

serious

no serious indirectness

no serious imprecision

none

250

257

-

MD -137.80 lower (-241.49 lower to -34.12 higher)

MD -137.80 lower (-241.49 lower to -34.12 higher)

IMPORTANT

 

Pneumonia rate

 

2

randomised trials

serious

serious

no serious indirectness

no serious imprecision

reporting bias

3/348
(0.9%)

3/361
(0.8%)

OR 1.04 (0.23 to 4.61)

0 more per 1000 (from 6 fewer to 29 more)

ÅOOO
VERY LOW

NOT IMPORTANT

 

 

0.9%

0 more per 1000 (from 7 fewer to 31 more)

 

Length of stay 

 

2

randomised trials

serious

no serious inconsistency

serious

no serious imprecision

reporting bias

348 

361

-

MD -0.65 lower (-1.32 lower to 0.02 higher)

ÅOOO
VERY LOW

NOT IMPORTANT