SciELO - Scientific Electronic Library Online

 
vol.37 número3Adding haloperidol to morphine for patient-controlled analgesia (PCA) reduces nausea and vomiting after short stay surgery: randomized, controlledEvaluating coagulation in prostatectomy índice de autoresíndice de assuntospesquisa de artigos
Home Pagelista alfabética de periódicos  

Serviços Personalizados

Journal

Artigo

Indicadores

Links relacionados

  • Em processo de indexaçãoCitado por Google
  • Não possue artigos similaresSimilares em SciELO
  • Em processo de indexaçãoSimilares em Google

Compartilhar


Colombian Journal of Anestesiology

versão impressa ISSN 0120-3347

Rev. colomb. anestesiol. v.37 n.3 Bogotá jul./set. 2009

 

Morbimortality in major hip surgery: A study of the efficacy of selective subarachnoideal spinal anaesthesia compared to balanced general anaesthesia. A controlled clinical trial

Héctor Julio Meléndez*, Arturo Mercado**, Julian Higuera Cobos***

Tercer puesto – Concurso Rafael Peña - XXVIII Congreso Colombiano de Anestesiología, marzo 2009, Bogotá.
* MD. Anestesiólogo Intensivista Epidemiólogo Profesor Asociado UIS. Email: hjmelendez@yahoo.com
** MD. Anestesiólogo UIS
*** MD. Anestesiólogo UIS - HUS

Recibido: febrero 18/2009 - Aceptado: julio 24/2009


SUMMARY

Objectives: Determining whether early mortality and at six months in elderly patients undergoing hip surgery was related to the type of anaesthesia used.

Methods: A controlled clinical trial was designed using 89 patients distributed into two groups. Patients were randomly assigned. Both techniques were standardised and all had an epidural catheter for 48 hrs POP. Morbidity was defined according to CIE X codification and mortality according to death certificate. Immediate post-operative follow-up was carried out in recovery and intrahospital periods until patients were discharged. Telephonic follow-up was continued after one month, the third month and after six months.

Results: Mortality was 14.61%, absolute and relative risks being greater in the general anaesthesia group of (20%) cf 9.09% having selective regional anaesthesia, but with no significant differences between them. Causes of death were cardiovascular (10.12%) and infectious aetiology (4.49%), having no differences between groups. 31.57% general morbidity was presented, having greater absolute and relative risks in the group intervened than in the control group, but without significant differences.

Conclusions: Early mortality and at six months with general anaesthesia compared to selective regional anaesthesia presented greater absolute and relative risks (20% cf 11.61% RR=2.2), having significant differences between both groups. Greater power is needed for demonstrating significance.

Key words: arthroplasty, replacement, hip, anesthesia, spinal, general, randomized controlled trial (source: MeSH, NLM).


INTRODUCTION AND OBJECTIVES

Hip fracture is an important health problem; it affects 80/100,000 inhabitants of the USA and it has been estimated that were 1.7 million fractures throughout the world in 1990. It has also been predicted for 2050 that this figure could reach 6.3 million around the world.(1) This problem could have negative repercussions on elderly adults´ quality of life of and on health-providing entities´ provision.

Most hip fractures are surgically managed, meaning that this population have become frequent users of anaesthesia services.(2) Given the high incidence of cardiovascular and other types of comorbidity presented in this population, there is an increased risk of morbimortality. Current knowledge and evidence do not lead to establishing clear differences in morbimortality regarding the type of anaesthesia administered to this type of patient.(3-8) The technique called selective subarachnoideal spinal anaesthesia (SSSA) was developed in our institution fourteen years ago (and as an ongoing line of research into anaesthesia) for lower limb surgery in low-risk patients.

It has been seen to be a technique which does not need prehydration for LEV, has excellent hemodynamic stability, requires minimum local anaesthetic dose, involves short recovery times and the evidence shows that pure selectivity can be achieved. This has led to the technique being extrapolated to high-risk patients and a series of cases being reviewed in which lower early mortality was found. The best form of evaluating this technique´s efficacy is by performing a controlled clinical trial (CCT).

MATERIALS AND METHODS

A CCT was carried out; it included ASA 1, 2, 3 elective and urgent patients as well as those without contraindications for being submitted to anaesthesia during surgery. Patients were excluded who had simultaneously undergone another surgical act or those who had been intervened during the previous 30 days, as well as patients suffering pathological fractures.

Sample size was calculated using a metanalysis database showing 9.44% average mortality with general anaesthesia (GA), seeking to reduce this by 25% by using SSSA with a=0.05 (95% significance) and β=0.2 (80% power). According to Fleiss´ formula, (9) 74 patients would be required. The resulting variables were morbidity and mortality at 30 days and at 6 months. Eligible patients and those fulfilling the inclusion criteria were randomly distributed into two groups: control group (GA) and intervention group (SSSA). Block randomisation was carried out for guaranteeing group size homogeneity. A random number table was generated, only allowing it to be known to which group a selected patient belonged immediately before he/she entered the operating room. Standard monitoring plus vesicle probe was applied to all patients and invasive monitoring was left up to the judgment of the anaesthesiologist responsible for managing a particular patient. Patients did not receive prior LEV or load and these were administered according to need, responding to loss from bleeding, fasting and surgery.

The lateral femoral cutaneous nerve had been previously blocked in the intervened group so that they could tolerate the procedure; later, in lateral position, the affected side was punctured at L3L4 or L2L3. Once there was evidence of LCR leaving, the 135 needle was rotated in caudal direction and 0.5% bupivacaine was applied without 7.5 mgrs epinephrine followed by 20 mcgrs of phentanyl for a three-minute period. The peridural catheter was then placed one level above the puncture and the patient was left for 10 minutes for fixing the anaesthetic and the technique´s selectivity. A peridural catheter for managing POP pain for 48 hours was placed in the control group prior to the anaesthetic induction or after a patient had become anaesthetised. The general balanced technique was left to the criteria of the anaesthesiologist treating the case; however, the drugs used were recorded on an instrument designed for such end. All patients submitted to SSSA received supplementary oxygen at 3lts /min by nasal cannula. Preventative analgesia was carried out with AINES in all patients. POP pain was managed by the Pain Clinic. Immediate POP follow-up was carried out during recovery and intrahospital period for all patients until they were discharged or they died. Follow-up was continued by telephone after one month, the third month and at six months.

The data recorded on the instrument was entered in duplicate in a database set up in Epiinfo 6.0 and validated by Epiinfo validate software. Both sets were compared; the differences found were reviewed until a single patient was left.

Their clinical characteristics were described using means and percentages and their respective of 95% confidence interval (CI). After verifying normal distribution, the data were analysed by using Fischer´s exact test and the t-test having 0.05 significance for evaluating differences between dichotomic and continuous variables, respectively. The risk of mortality and morbidity was estimated for each treatment group.(10) Their RR and 95% CI were used for mediating the efficacy of the intervention group compared to control. Univariate, bivariate and variance analysis were carried out for comparing groups. Logistical regression was also done to evaluate the independent effects of possible clinical risk factors for the presence of mortality and morbidity; p<0.05 was considered significant.(11) Cox´s logistical regression was used for analy sing survival. Goodness of fit was then applied to the final model. Stata 9.0 statistical software was used for all statistical analysis.

RESULTS

The present investigation was approved by the institution´s ethics committee and the informed consent form was signed by all patients. A total of 89 patients were recruited and randomly assigned as follows: 45 in the control group and 44 in the intervened group (SSSA).

There were no significant differences in base characteristics, except for ASA. Endoprothesis was practised on 24.89% of them (N=25). Average age was 77.43 and average body mass index (BMI) was 23.71 (Table 1).

No significant differences were presented regarding intraoperative (IOP) variables, except for time spent in surgery and the presence of cardiovascular instability due to hypotension and symptomatic bradycardia (Table 2).

Mortality was the main variable resulting from the present study. Minimum follow-up time was 36 days and maximum 468 days (176 average, 184 mean and 83 standard deviation). There were no losses during follow-up. Overall mortality incidence was 14.61%, absolute and relative risks being greater in the GA group (20% cf 9.09%) compared to selective regional anaesthesia (SRA) but having no significant differences between them (Table 3). The causes of death had cardiovascular (10.12%) and infectious aetiology (4.49%). There were no differences between the groups.

Bivariate and multivariate analysis for mortality

It was established whether there were a relationship between the presence of mortality (dependent variable) and each of the independent variables, including those having a biological meaning. Such relationship was evaluated in terms of relative risk and confidence interval. Vasopressor use, surgery time greater than 150 minutes, transfusion, CV IOP instability and pre-surgery wait greater than two days were risk factors, but without being significant. The other factors were protectors, in spite of presenting greater surgical risk, but without significant differences. These findings will lead to discussion (Table 4).

Logistical regression and modelling for mortality

Once it had been determined which independent variables were associated with the presence of mortality, a binomial logistical regression model was constructed by entering the “main” (type of anaesthesia) explanatory variable and all the rest which had been seen to have been associated with mortality (<0.20 statistical significance) or biological explanation. This final model showed that only the type of surgery (endoprothesis) was a protector factor, in spite of presenting greater relative risks for mortality. Having received GA, being older than 77, pre-surgery stay greater than 2 days and having presented CV IOP instability, none of the explanatory or independent or fuzzy variables behaved as independent factors or were significantly associated for mortality, meaning that irrespective of being risk factors or protectors, their statistical significance was not admitted in the final model. This finding will give rise to discussion (Table 5).

Analysis of survival at 180 days

Tables of life were thus constructed. Survival rate was evaluated by the Kaplan-Meier method and Cox´s proportional hazard model was used for obtaining mortality pattern regarding time and explanatory and fuzzy variables and evaluating the effect of each variable on mortality.

Analysing survival curves revealed that all mortality occurred during the first 73 days of follow-up, 50% of this during the first 35 days. Overall survival rate at 180 days for the intervened group was 88.52% (74.59– 95, 95%CI) and 77.73% (61.17-87.88, 95%CI) for the control group (Figure 1A and 1B). According to survival curve analysis, all mortality occurred during the first 3 months. It was thus proposed that this period of time should be evaluated, leading to the following results: total of patients during the period = 32, accumulative risk of death = 43.75%, number of deaths for every 1,000 cases /day observation = 7.91.

Survival curve analysis revealed that mean survival time was day 57 and the analysis by intervention group showed an earlier fall in mortality in the intervened group, having mean survival at day 36 compared to the control group where this occurred at day 64 (Figure 2).

Evaluating the other variables regarding the effect of the mortality in line with the Cox regression model, it was observed how none of the explanatory, independent and fuzzy variables were risk factors during the period of time being evaluated and the risk of mortality was equal to 3.15 for each day of follow-up (Table 6).

Morbidity

All the patients who died were excluded from morbidity analysis (N=13). There was a general morbidity rate of 31.57%, having greater absolute and relative risks in the intervened group than in the control group, but without significant differences. The causes of morbidity (grouped as infectious and cardiovascular) showed the predominance of both in the intervened group (10% and 20%) (Table 7).

The only variable in univariate and multivariate analysis showing significance for morbidity was being aged over 77 (3 RR; 1.34-6.72, 95%CI; p = 0.008) (Table 8).

DISCUSSION AND CONCLUSIONS

It should be stated for this discussion that 99% of the referents in the literature compare GA with classical regional anaesthesia (RA) (combined peridural-general anaesthesia) and not one controlled clinical trial was found in which the SRA technique was compared with the GA one regarding mortality.

No significant differences were found between the two anaesthetic techniques regarding early mortality and at 180 days. These results could have been similar in terms of absolute risks to the findings of Urwin et al., (3,12) who reported marginal advantage in favour of RA in their metanalysis, having less early mortality (one month). Findings regarding mortality were similar to those reported in most metanalysis, especially that of O´Hara et al., (5) who evaluated morbidity and mortality at 30 days in a population aged over 60 and found no association between type of anaesthesia and morbimortality. RA behaved as a protector factor for mortality (2.1% cf 3.1%) in the CORTRA study.(13) Other CCT, cohort and descriptive studies have failed in demonstrating differences between the two types of anaesthesia regarding mortality.(14-16)

No significant differences were found between both groups in our study regarding IOP bleeding, blood transfusions and LEV requirements; these did demonstrate differences in favour of the RA group in the metanalysis made by J. Mauermann et al., (17) as did that reported by Parker et al.,(2) who found greater blood loss and thromboembolic complications with GA. Surgery time presented significant differences in favour of the SSSA group, this being a very similar finding to the metanalysis of J. Mauermann et al.

POP analgesia has been one of the main advantages of RA and combined techniques; such difference was ignored in this study and all patients were given the same management with epidural catheter, which is why this difference did not exist in our study.

No significant differences were found between the two groups regarding POP morbidity, a differently to the metanalysis by J. Mauermann et al., (17) which concluded that RA offered better results than GA, regarding profound venous thrombosis, pulmonary thromboembolism and acute myocardial infarction, even though in the study of RA patients (whether epidural or subarachnoideal) they were involved in a single group.
No differences were found regarding POP cognitive dysfunction when analysing individual results, such findings being similar to those in the study by Casatii et al., (18) comparing SSSA to AG with sevorane. Most studies have not been conclusive regarding DCP and its relationship to anaesthetic technique (8-11), IOP events being more related to hypotension and hypoxemia.(1-5)

Regarding long-term results, our findings were very similar to those reported in world literature; although these may be scarce, no differences were found.(1)

The analysis of survival is worth noting as mortality occurred during the first 75 days POP and 50% of such mortality during the first 30 days whereas other studies reported findings at 30 and 180 days.

Debate will continue regarding the efficacy and safety of one of three anaesthetic techniques, not only in hip fractures but in all surgical procedures. The choice of anaesthetic technique does not seem to affect mortality in patients requiring surgery for hip fracture, in spite of persisting and finding greater opinion in favour of RA, even though this has not been supported by solid evidence.

Even though our working hypothesis was not confirmed, possibly due to a lack of power (sample size), it is hoped to carry out a second CCT and evaluate it again involving more patients.

REFERENCES

1. Zuckerman, JD. Hip fracture. N Engl J Med 1996; 334:1519-1525.

2. Parker, J Pryor G. Hip fracture management. Oxford: Blackwell scientific publication 2000.

3. Urwin S.C, Parker M.J, Griffitiths SR. General vs regional anesthesia for hip fracture surgery: a meta analysis of randomized trials. Br J anesth 2000.92. 947:957.

4. Breen, P., Kyung P. General anesthesia versus regional anesthesia. International Anesthesiology Clinics. 2002; 40(1): 61-71.

5. O´hara D.A Duff A , Berlin JÁ, Poses RM ,Lawrence VA Huber BEC Noveck H Strom BL Carson JL. The effect of anesthesic technique on postoperative outcomes in hip fracture repair, Anesthesiology 2000; 92:947-957.

6. Parker MJ, Handoll HG, Griffiths R. Anestesia para la cirugía de fractura de hip en adultos (translated from the Cochrane Review). In: the Biblioteca Cochrane Plus, número 2, 2005. Oxford, Update Software Ltd. Available at: http://www.update-software.com

7. Mckenzie P, Wishart Y, Smith G. Long-term outcome after repair of fracture neck of femur. Comparison of subarachnoid and general anesthesia. Br. Journal Anesth. 1984; 56(6): 581-5.

8. Dolk T. Hip fractures – Treatment and early complications. Ups J Med. Sci 1989; 94(2): 195-207.

9. Fleiss JL. Statistical methods for rates and proportions (2nd ed.) New York: Wiley-Interscience, 1981.

10. Rothman K, Rothman K, Greenland S., Lash T. Precision and Validity of studies. In Modern Epidemiology. Second edition 1998. Ed. Lippincott Williams & Wilkinscap. Buenos Aires Argentina. Ch.8, pp.119-133.

11. Greenland S. Introduction to Regression Modeling en Modern Epidemiology. Second edition 1998. Ed. Lippincott Williams & Wilkinscap. Buenos Aires Argentina. Ch.21, pp. 401-432.

12. Rodgers A, Walker N, Schug S, McKee A, Kehlet H, van Zundert A, et al. Reduction of postoperative mortality and morbidity anaesthesia: results from overview of randomised trials. Br Med J2000; 321:1493-1510.

13. The CORTRA study. A collaborative overview of randomized trials of regional anaesthesia. In: European Society of Regional Anaesthesia (ESRA) Annual Congress. Van Zundert A (editor) Publicity Permanyer, Barcelona, Spain. 1999. pp. 170-177.

14. Sorenson RM, Pace NL. Anesthetic technique during surgical repair of femoral neck fractures: a meta-analysis. Anesthesiology 1992; 77:1095-1104.

15. Koval KJ, Aharonoff GB, Rosenberg AD, Schmigelski C, Bernstein RL. Hip fracture in the elderly: the effect of anesthetic technique. Orthopedics 1999; 22:31-34.

16. Gilbert TB, Hawkes WG, Hebal JR, Hudson JI. Spinal anesthesia versus general anesthesia for hip fracture repair: a longitudinal observation of 741 elderly patients during 2-year follow up. Am J Orthoped 2000; 19:25-35.

17. Mauermann W, Shilling A., Zuo Zhiyi. A comparison of neuraxial block versus general anesthesia for elective total hip replacement: a meta-analysis (Anesth Analg 2006;103:1018 –25).

18. Casati, G Alldegheri, F Vinciguerra and et al. Randomized comparison between sevoflurane anaesthesia and unilateral spinal anaesthesia in elderly patients undegoing orthopaedic surgery. European Journal of Anaesthesiology 2003; 20; 640-646.

Conflicto de intereses: ninguno declarado.

1. Zuckerman, JD. Hip fracture. N Engl J Med 1996; 334:1519-1525.        [ Links ]

2. Parker, J Pryor G. Hip fracture management. Oxford: Blackwell scientific publication 2000.        [ Links ]

3. Urwin S.C, Parker M.J, Griffitiths SR. General vs regional anesthesia for hip fracture surgery: a meta analysis of randomized trials. Br J anesth 2000.92. 947:957.        [ Links ]

4. Breen, P., Kyung P. General anesthesia versus regional anesthesia. International Anesthesiology Clinics. 2002; 40(1): 61-71.        [ Links ]

5. O´hara D.A Duff A , Berlin JÁ, Poses RM ,Lawrence VA Huber BEC Noveck H Strom BL Carson JL. The effect of anesthesic technique on postoperative outcomes in hip fracture repair, Anesthesiology 2000; 92:947-957.        [ Links ]

6. Parker MJ, Handoll HG, Griffiths R. Anestesia para la cirugía de fractura de hip en adultos (translated from the Cochrane Review). In: the Biblioteca Cochrane Plus, número 2, 2005. Oxford, Update Software Ltd. Available at: http://www.update-software.com        [ Links ]

7. Mckenzie P, Wishart Y, Smith G. Long-term outcome after repair of fracture neck of femur. Comparison of subarachnoid and general anesthesia. Br. Journal Anesth. 1984; 56(6): 581-5.        [ Links ]

8. Dolk T. Hip fractures - Treatment and early complications. Ups J Med. Sci 1989; 94(2): 195-207.        [ Links ]

9. Fleiss JL. Statistical methods for rates and proportions (2nd ed.) New York: Wiley-Interscience, 1981.        [ Links ]

10. Rothman K, Rothman K, Greenland S., Lash T. Precision and Validity of studies. In Modern Epidemiology. Second edition 1998. Ed. Lippincott Williams & Wilkinscap. Buenos Aires Argentina. Ch.8, pp.119-133.        [ Links ]

11. Greenland S. Introduction to Regression Modeling en Modern Epidemiology. Second edition 1998. Ed. Lippincott Williams & Wilkinscap. Buenos Aires Argentina. Ch.21, pp. 401-432.        [ Links ]

12. Rodgers A, Walker N, Schug S, McKee A, Kehlet H, van Zundert A, et al. Reduction of postoperative mortality and morbidity anaesthesia: results from overview of randomised trials. Br Med J2000; 321:1493-1510.        [ Links ]

13. The CORTRA study. A collaborative overview of randomized trials of regional anaesthesia. In: European Society of Regional Anaesthesia (ESRA) Annual Congress. Van Zundert A (editor) Publicity Permanyer, Barcelona, Spain. 1999. pp. 170-177.        [ Links ]

14. Sorenson RM, Pace NL. Anesthetic technique during surgical repair of femoral neck fractures: a meta-analysis. Anesthesiology 1992; 77:1095-1104.        [ Links ]

15. Koval KJ, Aharonoff GB, Rosenberg AD, Schmigelski C, Bernstein RL. Hip fracture in the elderly: the effect of anesthetic technique. Orthopedics 1999; 22:31-34.        [ Links ]

16. Gilbert TB, Hawkes WG, Hebal JR, Hudson JI. Spinal anesthesia versus general anesthesia for hip fracture repair: a longitudinal observation of 741 elderly patients during 2-year follow up. Am J Orthoped 2000; 19:25-35.        [ Links ]

17. Mauermann W, Shilling A., Zuo Zhiyi. A comparison of neuraxial block versus general anesthesia for elective total hip replacement: a meta-analysis (Anesth Analg 2006;103:1018 -25).        [ Links ]

18. Casati, G Alldegheri, F Vinciguerra and et al. Randomized comparison between sevoflurane anaesthesia and unilateral spinal anaesthesia in elderly patients undegoing orthopaedic surgery. European Journal of Anaesthesiology 2003; 20; 640-646.        [ Links ]