Outcomes and cost-effectiveness of different anesthesia techniques in elderly patients with hip fracture

Abstract

AimThis study aimed to determine the optimal type of anesthesia for elderly patients to reduce the need for intensive care, mortality, and costs of operations for hip fractures.MethodsIn this retrospective and cross-sectional study, the records of 204 patients aged 60 years and older who were operated on for hip fracture with neuraxial (107 patients) or general anesthesia (97 patients) at Pamukkale University Hospital. The duration of intensive care stay, duration of hospital stay, mortality, and cost were considered primary endpoints.ResultsAll evaluation parameters were evenly distributed between both groups. Among the analyzed parameters, the duration of surgery and the rate of patients requiring >1 unit of blood transfusion were found to be significantly lower in neuraxial anesthesia patients, compared to general anesthesia patients (P < .001,001). Intensive care need, duration of intensive care stay and duration of hospital stay, three-month mortality, and cost were found to be significantly lower in the neuraxial anesthesia group, compared to the general anesthesia group (for all P = .001). Based on logistic regression analysis, age (age 80-89, OR= 2.782, 95% CI= 1.156-6.694; age 90 +, OR= 8.779, CI= 2.056-37.486), general anesthesia (OR = 8.069, 95% CI= 2.778-23.436) and need of intensive care (OR= 5.155, 95% CI= 2.233-11.903) were independent risk factors affecting mortality.ConclusionThe results of our study show that the use of neuraxial anesthesia in geriatric hip fracture surgery has significant advantages, including lower costs and mortality rates compared to general anesthesia.

Keywords

Introduction

Hip fracture is a frequent problem in the elderly population. The predicted number of hip fractures by 2050 is 6.3 million.¹ Approximately 8.4% of patients die within 30 days after hip fracture surgery and 15-30% in the following year.² The effect of anesthesia techniques used in elderly patients on outcomes has not yet been clarified in the literature. Available scientific evidence that will enable the selection of the most appropriate anesthesia technique in femur fracture surgery to be used is still limited.²
The costs related to intensive care services are approximately 22% of all hospital expenditures, while intensive care beds constitute only 10% of all hospital beds.³ Thus, intensive care expenses have an important role in the total cost of surgery-related hospitalizations such as hip fracture treatments. Although the use of neuraxial anesthesia in hip fracture surgery is shown to result in a reduction in intensive care stay and related costs, the multifactorial reasons for this advantage are not yet all elucidated.
Our primary aim in this study is to compare the impact of anesthesia techniques in hip fracture surgery on postoperative intensive care admission, mortality, and cost. Our hypothesis is that local anesthesia is superior to general anesthesia in terms of specified criteria in elderly hip fracture patients. We also wanted to determine the independent effects of variables of the patient and treatment process on mortality.

Materials and Methods

In this retrospective cross-sectional study, hospital information management system records and anesthesia intensive care files of 287 patients aged 60 years and older who were operated on with neuraxial or general anesthesia due to hip fracture at Pamukkale University Medical Faculty Hospital were retrospectively analyzed. The choice of anesthetic technique in our study, i.e. general or neuraxial, depended on factors related to both patient consent and the consultation of the anesthesiologist. Both procedures were explained to the patients in detail with advantages and disadvantages, and patient consent was obtained.
Patients were placed in either general anesthesia (GA) or neuraxial anesthesia (NA) group, which included either spinal or epidural anesthesia patients. Patients who met the exclusion criteria were excluded from the study (Figure 1).
Information such as demographic data, history of smoking, ASA risk assessment, Charlson Comorbidity Index (CCI).⁴ preoperative laboratory parameters, type of surgery, type of fracture, preoperative intensive care need, duration of operation, use of cement, and amount of intraoperative blood transfusion were obtained from the hospital records. Length of stay in the intensive care unit, duration of hospital stay, mortality, and cost were considered primary endpoints. Mortality rates were assessed at 7 days, 1 month, and 3 months.
Preoperative laboratory parameters were taken from the last lab reports before surgery. Costs were calculated according to the average US dollar rates at the hospitalization date of patients. The cost calculation was made according to the duration of intensive care and hospital stay. Implant costs were subtracted from the total cost because of the variability in the type and cost of implants.Ethical ApprovalThis study was approved by the Ethics Committee of Pamukkale University, Faculty of Medicine (Date: 27.03.2013, Decision No: 60116787/64).Statistical AnalysisFor the statistical analysis of data, the Statistical Package for the Social Sciences (SPSS) 18.0 software was used. The results were assessed with a 95% confidence interval at P = .05 significance level.Reporting GuidelinesThis study was reported in accordance with the STROBE guideline.

Results

Out of the 204 patients who met the study criteria, 107 received GA, and 97 NA. The selection of patients is shown in the flow diagram (Figure 1). There were no statistically significant differences between the GA and NA groups, regarding the distribution of age, sex, history of smoking, ASA, CCI, and preoperative laboratory parameters.
The duration of surgery and the rate of patients who needed more than one unit of blood transfusion were found to be statistically significantly lower in the NA patients. Transfer to the intensive care, duration of intensive care, hospital stay, one-month and three-month mortality, and cost (USD) were found to be significantly lower in the NA group (Table 1). Since the ICU length of stay variable was skewed to the right, the analysis was conducted by excluding outliers. Among 107 patients who received GA, 70 were predicted to be transferred to the intensive care unit, and 53 of them ultimately were transferred (53/70, 76%); while of the 97 patients who received NA, 53 were predicted to be transferred to the intensive care, but only 18 were eventually transferred (18/53, 34%) (P < .05). The evaluation of all cases showed that, in cases who were transferred to the intensive care unit, some parameters such as ASA III score (I- II or III, 67.6% vs 33.8%), CCI >3 score (<3 or >3, 59.2% vs 26.3%), and need for more than 1 unit of blood transfusion (63.4% vs 27.1%) were found to be statistically significantly higher, preoperative albumin rates were significantly lower and duration of surgery was longer compared to patients who did not go to the intensive care unit (P < .05). The evaluation of risk factors affecting mortality showed that some variables, which were considered significant in the univariate analysis (Table 2) were insignificant in the logistic regression analysis. It was confirmed by Collinearity Statistics that there were no multiple linear linkages between independent variables to be included in logistic regression. Age groups, ASA, CCI, type of anesthesia, blood transfusion and intensive care stay variables were included in the Forward Stepwise (Conditional) model. As a result of the analysis, age, GA and need for intensive care were assessed as independent variables for mortality (Table 3). The most common cause of mortality was heart failure. Clinically, it was thought that the ‘entrance to ICU’ variable might have a mediator role in the effect of ‘Anesthesia type’ on mortality, and mediation analysis was performed. Other variables were excluded from this analysis because they were eliminated in the logistic regression model and the ‘age group’ variable had no effect on the type of anesthesia. According to the mediation analysis results, the independent variable “anesthesia type” (NA) affects mortality significantly and negatively (β = -2.233 P < .001 95% CI = 0.040-0.287). ‘Anesthesia type’ (NA) affects the mediator variable “entrance to ICU” statistically significantly and negatively (β = -1.460 P < .001 95% CI = 0.123-0.439). Similarly, entrance to ICU affects mortality (3 months) statistically significantly and positively (β = 1.941 P < .001 95% CI = 3.260-14.870). In the last stage of the mediation analysis, it was observed that the effect of anesthesia type on mortality was still significant as in the first model when the entrance to ICU variable was included in the model, but there was a significant decrease in the Beta coefficient (= -1.851 P < .001 95% CI = 0.057-0.435) (β -2.233 to -1.851). These findings show that the “entrance to ICU” variable has a partial mediator effect in the relationship between anesthesia type and mortality.

Discussion

Hip fracture surgery is a common medical procedure in elderly patients, which can result in high morbidity and mortality due to factors related to the patient and/or surgery.⁵ Perioperative mortality and morbidity are affected by the choice of anesthesia method, concomitant diseases, and surgical technique.⁶
Anesthesia techniques used in hip fracture surgery have often been compared previously, however, studies that evaluated different anesthesia techniques in terms of both patient results and cost-effectiveness were lacking. Our study focuses on this comparison.
In the preoperative phase, 53/97 patients who received NA were predicted to be in need of intensive care, while eventually, only 18 needed it. Admission to intensive care, duration of intensive care stay, and duration of hospital stay were found to be significantly lower in the NA group compared to the GA group. Similarly, Kaufmann et al. stated that NA can reduce the postoperative intensive care requirements in high-risk patients who are to undergo hip prosthesis surgery. They also reported that NA reduces the duration of intensive care and hospital stay, and therefore reduces cost.³ Ahn et al. also found that NA was associated with better outcomes than GA in terms of ICU admission and ventilator care in elderly patients undergoing hip fracture surgery.⁷ There are other studies showing that regional anesthesia is associated with a shorter hospital stay compared to GA.^8-10
The decision of anesthesiologists or surgeons on the postoperative destination of these patients probably depends on the initiative of the caregiving physician because of the inadequacy of standard anesthesia intensive care admission guidelines. This relatively subjective decision may be based on the estimation of how well the patient will tolerate the surgical procedure. Patients who receive NA may seem more stable during the operation process, and this can affect the decision of the clinician regarding transfer to the intensive care unit. Additionally, intraoperative blood loss has been reported to be significantly lower in patients who receive NA, therefore, the need for blood transfusion, and probably fluid replacement is also relatively low.^11-13 Since NA has various intraoperative benefits that will affect the general physical status of the patient after surgery, this may affect the decision of the clinician on the need for intensive care.
Our study showed that there was less need for blood transfusion in the NA group compared to the GA group (Table 2). The recent change in surgical methods and the increase in the rate of closed surgery may have been another factor reducing the need for blood transfusion in our study. Various studies have investigated the relationship between anesthesia techniques and intraoperative blood loss.^1,14 NA was reported to be related to less bleeding in patients with hip fracture surgery.¹⁴ Venous pooling due to sympathetic blockage, vasodilation, and neuraxial blockage reduces the venous return to the heart. However, Urwin et al. showed in their meta-analysis of the data from 3 studies that there was no difference between general and NA in terms of blood transfusion.¹
Hip fractures in the geriatric population create an economic burden on the health system. In our study, a significant reduction was found in the use of intensive care, duration of intensive care stay, and duration of hospital stay, and therefore hospital costs, with the use of NA (Table 2). It has been suggested in a study that length of stay and surgical choice were linearly related to total hospital costs.¹⁵ There are few studies on the economic burden of the type of anesthesia used in intensive care patients. There are some studies showing that neuraxial anesthesia can improve pulmonary outcomes and reduce resources (critical care services, hospital stay) compared to general anesthesia.^3,16 In a study in England, in 2000, the cost per hip fracture patient was calculated as 4760 pounds, whereas in 2008, this number was calculated as 8000 pounds.¹⁷
Our logistic regression analysis showed that older age is an independent risk factor for mortality. Similarly, Mc Leod et al. have reported that the increase in systemic concomitant diseases in older age has an impact on mortality.¹¹
We found that high CCI risk scores, which indicate multiple medical problems, were associated with mortality; 29.8% of the patients with a CCI score above 3 died. However, the CCI score was not found to be an independent risk factor in our logistic regression analysis (Table 3). Souza et al. investigated the use of CCI as a method for estimating 90-day mortality risk in elderly patients hospitalized for hip fracture, and they showed that as the CCI score increased, the mortality risk also increased.¹³
In our study, 3-month mortality in ASA III patients (26%) was higher than ASA I and II patients (14%) among the 204 patients. While the difference between these rates was significantly high, ASA physical status scale was not determined to be an independent risk factor for mortality, based on multivariate statistical analysis. Holvik et al. found that those who died within 1 year had a higher ASA score (III, IV) and therefore more comorbid conditions.¹⁸ In this study, the 3-month mortality (short-term mortality) was 25.2% in 107 patients in the GA group. On the other hand, it was found to be significantly lower in the NA group (3.2%). Logistic regression analysis showed that the mortality risk was found to be 8.07 times higher in the GA group compared to the NA group and 5.16 times higher in the group admitted to the ICU. In mediation analysis, it was seen that the effect of “anesthesia type” on mortality was partially through “admittance to intensive care”. Therefore, this situation should be taken into consideration when evaluating odds ratios.
A limited number of studies have reported that postoperative mortality in NA patients is lower than in GA patients.^7,12 Desai et al. found that the use of general anesthesia was associated with a higher risk of mortality during the in-hospital stay compared with regional anesthesia, but this higher risk did not persist after hospital discharge.¹⁹ Chen et al. also reported a similar result, general anesthesia was associated with an increased risk of in-hospital mortality.⁹ However, many later studies have shown that, conversely, the type of anesthesia does not affect postoperative mortality.^8,10,16
Gilbert et al. found no significant difference between patients who received spinal or GA in long-term (two-year) mortality and incidence of serious morbidity, in a study with 741 patients operated for acute hip fracture.²⁰ Katušin et al. showed that the mode of anesthesia (general and spinal) had no effect on postoperative mortality and suggested that the patient’s comorbidities should be considered in the selection of anesthesia.²¹ Parker et al. found in their study that after NA, a 1-month mortality decrease is the limit.¹⁷ Hospital mortality (up to 1 month) and short-term mortality (up to 3 months) are lower with spinal anesthesia, but this advantage disappears after 3 months.²¹ Rodgers et al. have reported that mortality decreased by 30% with NA in major surgeries, including hip fractures.²²

Limitations

This study has some limitations due to its retrospective design. In addition, the initial design and set-up of the dataset limited the use of ordinal data. Examples of this are the analysis of the Charlson Comorbidity Index values and the need for blood transfusion, both of which are evaluated as binary rather than ordinal data while creating the dataset from the patient files. This approach may have reduced the sensitivity of the analysis performed with these two variables.

Conclusion

Overall, the results of our study show that the use of NA in geriatric hip fracture surgery is related to a lower need for blood transfusion, treatment costs, shorter intensive care and hospital stay, and a reduced rate of intensive care necessity. Neuraxial anesthesia was found to reduce mortality by 8.06 times compared to GA. Our findings support the recommendations of the International Consensus on Anesthesia-Related Outcomes after Surgery group as stated “neuraxial anesthesia is recommended for hip arthroplasty given associated outcome benefits; evidence level: moderate-low, strong recommendation”.²³
Longer-term prospective prognosis and cost analysis studies should be conducted to establish criteria for the selection of the appropriate anesthesia type in hip fracture surgery in elderly patients.

Declarations

Abbreviations

References

1. Urwin SC, Parker MJ, Griffiths R. General versus regional anaesthesia for hip fracture surgery: a meta-analysis of randomized trials. Br J Anaesth. 2000;84(4):450-455. doi:10.1093/oxfordjournals.bja.a013468
   Article · PubMed · Google Scholar
2. Membership of the Working Party, Griffiths R, Alper J. Management of proximal femoral fractures 2011: Association of Anaesthetists of Great Britain and Ireland. Anaesthesia. 2012;67(1):85-98. doi:10.1111/j.1365-2044.2011.06957.x
   Article · PubMed · Google Scholar
3. Kaufmann SC, Wu CL, Pronovost PJ, Jermyn RM, Fleisher LA. The association of intraoperative neuraxial anesthesia on anticipated admission to the intensive care unit. J Clin Anesth. 2002;14(6):432-436. doi:10.1016/s0952-8180(02)00392-6
   Article · PubMed · Google Scholar
4. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40(5):373-383. doi:10.1016/0021-9681(87)90171-8
   Article · PubMed · Google Scholar
5. Mackey DC. Physiological effects of regional block. In: Brown DL, ed. Regional anesthesia and analgesia. 1st ed. WB Saunders; 1996:397-422.
   PubMed · Google Scholar
6. Halachmi S, Katz Y, Meretyk S, Barak M. Perioperative morbidity and mortality in 80 years and older undergoing elective urology surgery: a prospective study. Aging Male. 2008;11(4):162-166. doi:10.1080/13685530802351081
   Article · PubMed · Google Scholar
7. Ahn EJ, Kim HJ, Kim KW, Choi HR, Kang H, Bang SR. Comparison of general anaesthesia and regional anaesthesia in terms of mortality and complications in elderly patients with hip fracture: a nationwide population-based study. BMJ Open. 2019;9(9). doi:10.1136/bmjopen-2019-029245
   Article · PubMed · Google Scholar
8. Neuman MD, Rosenbaum PR, Ludwig JM, Zubizarreta JR, Silber JH. Anesthesia technique, mortality, and length of stay after hip fracture surgery. JAMA. 2014;311(24):2508-2517. doi:10.1001/jama.2014.6499
   Article · PubMed · Google Scholar
9. Chen DX, Yang L, Ding L, Li SY, Qi YN, Li Q. Perioperative outcomes in geriatric patients undergoing hip fracture surgery with different anesthesia techniques: a systematic review and meta-analysis. Medicine (Baltimore). 2019;98(49). doi:10.1097/md.0000000000018220
   Article · PubMed · Google Scholar
10. Nishi T, Maeda T, Imatoh T, Babazono A. Comparison of regional with general anesthesia on mortality and perioperative length of stay in older patients after hip fracture surgery. Int J Qual Health Care. 2019;31(9):669-675.
    PubMed · Google Scholar
11. McLeod K, Brodie MP, Fahey PP, Gray. Long-term survival of surgically treated hip fracture in an Australian regional hospital. Anaesth Intensive Care. 2005;33(6):749-755. doi:10.1177/0310057x0503300608
    Article · PubMed · Google Scholar
12. Meyer HE, Tverdal A, Falch JA, Pedersen JI. Factors associated with mortality after hip fracture. Osteoporos Int. 2000;11(3):228-232. doi:10.1007/s001980050285
    Article · PubMed · Google Scholar
13. De Souza RC, Pinheiro RS, Coeli CM, de Camargo KR. The Charlson comorbidity index for adjustment of hip fracture mortality in the elderly: analysis of the importance of recording secondary diagnoses. Cad Saude Publica. 2008;24(2):315-322. doi:10.1590/s0102-311x2008000200010
    Article · PubMed · Google Scholar
14. Guay J, Parker MJ, Gajendragadkar PR, Kopp S. Anaesthesia for hip fracture surgery in adults. Cochrane Database Syst Rev. 2016;(2). doi:10.1002/14651858.cd000521.pub3
    Article · PubMed · Google Scholar
15. Hardy KM, Kwong J, Pitzul KB, et al. A cost comparison of laparoscopic and open colon surgery in a publicly funded academic institution. Surg Endosc. 2014;28(4):1213-1222. doi:10.1007/s00464-013-3311-y
    Article · PubMed · Google Scholar
16. Smith LM, Cozowicz C, Uda Y, Memtsoudis SG, Barrington MJ. Neuraxial and combined neuraxial/general anesthesia compared to general anesthesia for major truncal and lower limb surgery: a systematic review and meta-analysis. Anesth Analg. 2017;125(6):1931-1945. doi:10.1213/ane.0000000000002069
    Article · PubMed · Google Scholar
17. Chakladar A, White SM. Cost estimates of spinal versus general anaesthesia for fractured neck of femur surgery. Anaesthesia. 2010;65(8):810-814. doi:10.1111/j.1365-2044.2010.06382.x
    Article · PubMed · Google Scholar
18. Holvik K, Ranhoff AH, Martinsen MI, Solheim LF. Predictors of mortality in older hip fracture inpatients admitted to an orthogeriatric unit in Oslo, Norway. J Aging Health. 2010;22(8):1114-1131. doi:10.1177/0898264310378040
    Article · PubMed · Google Scholar
19. Desai V, Chan PH, Prentice HA, et al. Is anesthesia technique associated with a higher risk of mortality or complications within 90 days of surgery for geriatric patients with hip fractures? Clin Orthop Relat Res. 2018;476(6):1178-1188. doi:10.1007/s11999.0000000000000147
    Article · PubMed · Google Scholar
20. Gilbert TB, Hawkes WG, Hebel JR, et al. Spinal anesthesia versus general anesthesia for hip fracture repair: a longitudinal observation of 741 elderly patients during 2-year follow-up. Am J Orthop (Belle Mead NJ). 2000;29(1):25-35.
    PubMed · Google Scholar
21. Lončarić-Katušin M, Mišković P, Lavrnja-Skolan V, Katušin J, Bakota B, Žunić J. General versus spinal anaesthesia in proximal femoral fracture surgery: treatment outcomes. Injury. 2017;48(suppl 5):51-55. doi:10.1016/s0020-1383(17)30740-4
    Article · PubMed · Google Scholar
22. Rodgers A, Walker N, Schug S, et al. Reduction of postoperative mortality and morbidity with epidural or spinal anaesthesia: results for overview of randomised trials. BMJ. 2000;321(7275):1493. doi:10.1136/bmj.321.7275.1493
    Article · PubMed · Google Scholar
23. Memtsoudis SG, Cozowicz C, Bekeris J, et al. Anaesthetic care of patients undergoing primary hip and knee arthroplasty: consensus recommendations from the International Consensus on Anaesthesia-Related Outcomes after Surgery group based on a systematic review and meta-analysis. Br J Anaesth. 2019;123(3):269-287. doi:10.1016/j.bja.2019.05.042
    Article · PubMed · Google Scholar

Tables