Print version ISSN 0034-7094
Rev. Bras. Anestesiol. vol.57 no.6 Campinas Nov./Dec. 2007
Association between fasting blood glucose levels and perioperative morbimortality: retrospective study in surgical elderly patients*
Asociación entre glicemia de ayuno y morbimortalidad perioperatoria: estudio retrospectivo en pacientes ancianos quirúrgicos
Arthur Vitor Rosenti Segurado, TSAI; Flavia Salles Souza Pinotti Pedro, TSAII; Judymara Lauzi Gozzani, TSAIII; Lígia Andrade da Silva Telles Mathias, TSAIV
Assistente, Serviço Médico de Anestesia; Mestre em Medicina pela
Faculdade de Ciências Médicas da ISCMSP
IIMédica Assistente, Serviço de Anestesia da ISCMSP; Mestre em Medicina pela Faculdade de Ciências Médicas da ISCMSP
IIIMestre em Biologia Molecular; Doutora em Medicina; Responsável pelo Grupo de Dor da ISCMSP
IVDiretora do Serviço e Disciplina de Anestesiologia, ISCMSP e Faculdade de Ciências Médicas da ISCMSP; Responsável pelo CET/SBA da ISCMSP
OBJECTIVES: The relationship between altered blood glucose levels and perioperative
complications in elderly patients undergoing surgeries are not known. The objective
of this study was to evaluate the association between fasting blood glucose
levels and perioperative morbimortality in elderly surgical patients.
METHODS: Medical records of patients older than 60 years undergoing several surgical procedures during a 6-month period were analyzed and divided, according to fasting blood glucose levels, in three groups: < 100 mg.dL-1, between 100 and 125 mg.dL-1, and > 126 mg.dL-1. Age, physical status (ASA), history of diabetes mellitus (DM), and treatment and perioperative surgery-specific cardiologic risk were analyzed. Using univariate analysis and a model of multivariate logistic regression, the relationship among the outcome (frequency of postoperative complications [POC] and death) and the following variables: fasting blood glucose, history of DM, physical status (ASA), and cardiac risk, was evaluated.
RESULTS: A statistical association was demonstrated only among the three groups and physical status and history of DM. All parameters studied demonstrated a statistical relationship regarding the higher frequency of POC and death, except for the parameter history of DM, which did not demonstrate any relationship with deaths. In the multivariate logistic regression model, there was an association between cardiac risk and blood glucose levels with POC, while only physical status and cardiac risk demonstrated a statistical association with death.
CONCLUSIONS: This retrospective study demonstrated a significant association among blood glucose levels above 100 mg.dL-1 and postoperative morbimortality in the elderly.
Key Words: COMPLEMENTARY LABORATORY TESTS: blood glucose levels; COMPLICATIONS: morbidity, mortality; DISEASES: diabetes mellitus; SURGERY, Geriatric.
Y OBJETIVOS: Las relaciones entre valores alterados de glicemia y complicaciones
peri operatorias en la población de ancianos sometidos a procedimientos
quirúrgicos todavía no se conocen. El objetivo de este estudio
fue evaluar la Asociación entre glicemia de ayuno y morbimortalidad perioperatoria
en pacientes quirúrgicos ancianos.
MÉTODO: Se analizaron las hojas clínicas de pacientes de más de 60 años sometidos a diversos procedimientos quirúrgicos en un período de 6 meses, divididos de acuerdo con los valores de glicemia de ayuno en los grupos: < 100 mg.dL-1, entre 100 y 125 mg.dL-1 y > 126 mg.dL-1. Fueron analizados en cuanto a la edad, estado físico (ASA), historial previo de diabetes melito (DM) y tratamiento y riesgo cardíaco perioperatorio cirugía específica. A través de análisis univariada y de un modelo de regresión logística multivariada, se evaluó la relación entre los desenlaces (frecuencia de complicaciones postoperatorias (CPO) y óbitos) y las variables: glicemia de ayuno, historial previo de DM, estado físico (ASA) y riesgo cardíaco.
RESULTADOS: Hubo asociación estadística apenas entre los grupos glicémicos y las variables estado físico e historial previo de DM. Todas las variables estudiadas revelaron asociación estadística con relación a la mayor frecuencia de CPO y óbitos, excepto la variable presencia de historial previo de DM que no presentó relación con óbitos. En el modelo de regresión logística multivariada, hubo asociación entre las variables riesgo cardíaco y glicemia en función de la incidencia de CPO, mientras apenas las variables estado físico y riesgo cardíaco revelaron asociación estadística en función de los óbitos.
CONCLUSIONES: Ese estudio retrospectivo mostró que, para la población pacientes ancianos estudiada, hubo asociación significativa entre glicemias por encima de 100 mg.dL-1 y morbimortalidad perioperatoria.
Surgeries in the elderly are related with increased morbimortality, possibly due to comorbidities associated with this population, reduced physiologic reserve in the elderly, or both 1,2.
The increased prevalence of several comorbidities in one individual hinders stratification of patients in smaller, risk-defined groups. Besides, the scarcity of relevant studies on perioperative risks specific for the elderly, and the unique characteristics of this population make it difficult to obtain reliable data and to reach definitive conclusions. For those reasons, despite the recognition that diabetes mellitus (DM) is associated with increased perioperative morbidity in the general population, data regarding the elderly have not brought a consensus on to the extension this disorder contributes specifically to promote adverse events in the perioperative period 3.
Diabetes mellitus does not seem to be associated only with perioperative adverse events. There is evidence that stress-induced hyperglycemia (SIH) is also related with increased perioperative morbimortality 4.
The association between hyperglycemia and increased morbimortality is already known in specific medical situations, such as patients who suffered cerebral vascular accident 5,6, head trauma 7, myocardial infarction 8, or in severely ill patients who need intensive care 9.
Rigorous control of blood glucose levels by continuous administration of insulin in hyperglycemic patients reduced significantly the incidence of sternal infections after cardiac surgeries and decreased the in-hospital morbidity in patients undergoing surgeries who needed intensive care, suggesting that hyperglycemia might be responsible for the increased perioperative morbidity 10-12.
Despite the evidence, hyperglycemia is not included in the predictive indexes of cardiac risk or perioperative complications. Besides, there is a lack of scientific studies on hyperglycemia and perioperative complications in the elderly.
Questions regarding this subject motivated the present study, whose objective was to evaluate the association between fasting blood glucose levels and perioperative morbimortality in surgical, elderly patients by the analyzing fasting blood glucose levels and their relationship with parameters potentially associated with postoperative complications and death.
After approval by the Ethics Committee for Research of the Hospital Central da Irmandade da Santa Casa de Misericórdia de São Paulo, this retrospective study consisted of gathering data from medical records of elderly patients (aged 60 or older) who underwent different types of surgeries under anesthesia during a 6-month period. Exclusion criteria included: < 60 years old; local anesthesia; information hard to read in the charts; or incomplete records.
Data gathered included: gender; age; blood glucose levels; physical status according to the ASA classification; history of DM and treatment; perioperative surgery-specific cardiac risk (PSS); postoperative complications (POC); and deaths.
After selecting the medical records that would be included in the study, the data was recorded on standardized forms.
A descriptive analysis of the data was undertaken. Patients were divided in three groups according to blood glucose levels: < 100mg.dL-1; 100-125 mg.dL-1; and > 126 mg.dL-1, according to the diagnostic and classification criteria of DM of the American Diabetes Association 13. ANOVA was used to compare results regarding age. The Chi-square (c2) test was used to compare the other variables, and the level of significance was established at 0.05 (a = 50%). Univariate analysis was used for the following parameters: physical status (ASA), PSS cardiac risk, blood glucose levels, and history of DM in relation to the outcomes of postoperative complications and deaths, and analyzing the impact of those parameters by assembling a logistic multivariate regression model. The results of the logistic regression model are expressed in Odds Ratio. The test based on the Maximum Probability Ratio was used for the statistical analysis of the adjustment of the model. The 95% confidence interval was calculated for the estimate produced. The tests used are part of the statistical package Sigma Stat for Windows, version 2.03, SPSS Inc.
Out of 502 medical records analyzed, only 438 had blood glucose levels recorded, and that was considered the total size of the sample. The maximal, minimal, and mean values and standard deviation for blood glucose levels (mg.dL-1) were, respectively: 500.0; 61.0; 119.4; and 51.6.
Patient distribution showed that 39.5% were females and 60.5% males.
Patient distribution according blood glucose levels was as follows: < 100 mg.dL-1, n = 180 (41.1%); 100-125 mg.dL-1, n = 148 (33.8%); > 126 mg.dL-1, n = 110 (25.1%).
Mean values and standard deviation for age in all three groups were: < 100 mg.dL-1: 72.4 ± 7.6; 100-125 mg,dL-1: 74.2 ± 8.7; > 126 mg.dL-1: 73.2 ± 6.9. Groups were not statistically different (ANOVA p = 0.122).
Analysis of groups with different levels of blood glucose regarding physical status (ASA) indicates a greater number of patients ASA II in all three groups (Table I), and the frequency of patients ASA III and IV was almost four fold greater in the group with blood glucose level > 126 (c2 p < 0.001). Physical status ASA III and IV were analyzed together due to the small number of patients ASA IV in the study.
Statistical analysis of the distribution of a history of DM in all three groups revealed statistically significant differences (c2 p < 0.001) (Table II).
The frequency of treatment for DM referred in the charts was also analyzed: 53 (46.7%) patients of those with a past history of DM, and 3 (0.9%) patients without history of DM reported being treated for this condition (c2 p < 0.0001). Treatment included insulin and/or oral hypoglycemic drugs.
Table III shows the frequency of low, intermediate, and high PSS cardiac risk in all three groups, which did not demonstrated statistically significant differences (c2 p = 0.18).
Sixty-nine patients (15.8%) had some postoperative complication and 26 patients (5.9%) died.
Table IV shows the results of the univariate analysis of the relationship among different parameters: physical status (ASA), PSS cardiac risk, different blood glucose levels, history of DM and frequency of postoperative complications, and the significance of the statistical tests used.
Table V shows the results of the multivariate analysis, defining the parameters that remained in the model. For the outcome of POC they were blood glucose level and PSS cardiac risk, and for deaths they were physical status and PSS cardiac risk.
The impact of each parameter on the outcome was analyzed using the multivariate logistic regression model.
Table VI shows the multivariate logistic regression model when assessing the influence of the study parameters (PSS cardiac risk and blood glucose level) against postoperative complications.
Table VII shows the multivariate logistic regression model in the assessment of physical status and PSS cardiac risk against the incidence of deaths.
Since this was a retrospective study, it has some limitations because the data recorded in the medical records did not follow uniform criteria, being subjected to human and communication failures 14.
The comparison of physical status (ASA) and groups with different blood glucose levels showed higher frequency of ASA III and IV and smaller frequency of ASA I patients with blood glucose levels higher than 126 mg.dL-1. This result was expected, since patients ASA III and IV have higher incidence of comorbidities than patients ASA I and II, and some of them might be related to elevated blood glucose levels, such as DM. It is interesting to note that 14 out of 107 patients with a history of DM (13.1%) had fasting blood glucose levels below 100 mg.dL-1. Seventy out of 107 patients with positive history of DM (65.4%) had blood glucose levels greater than 126 mg.dL-1 demonstrating, therefore, to be decompensated. Among all patients with a history of DM, less than half (46.7%) referred some form of treatment (diet, oral hypoglycemic agents, insulin, or a combination), suggesting poor adherence to treatment. Three patients without history of DM referred treatment of this disease. This might be due to misinformation of patients regarding their past medical history, medications they were taking or mistakes made when filling out the forms in the charts of the patients.
One could observe that out of 110 patients with blood glucose levels above 126 mg.dL-1, 40 (36.4%) did not have a history of DM. Despite recommendations that the diagnosis of DM should not be made with only one blood glucose level, those patients had increased probability of being diabetics without prior knowledge 15. Other studies, especially those from patients with history of a recent myocardial infarction, had elevated indexes of undiagnosed DM. Oswald et al. 16 and Tenerz et al. 17 suggested that this prevalence would be up to 4%. Bolk et al. 18 found that 8.1% of individuals with myocardial infarction without a history of DM had a blood glucose level on admission above 200 mg.dL-1. These results demonstrate that DM is frequently diagnosed due to a blood glucose level that was requested as part of the investigation of other clinical situations, which agrees with statistics showing that a large proportion of the diagnosis of DM are done due to routine preoperative laboratorial exams, without prior symptoms 19. These considerations are pertinent since DM can cause late symptoms, in the presence of significant metabolic imbalance or as the manifestation of some diabetes mellitus-related disease (for example, myocardial infarction) 8. These data reinforce the importance of determining blood glucose levels in the population studied, since the preoperative identification of hyperglycemic patients offers the opportunity for an early and adequate therapeutic approach.
Patient distribution regarding PSS cardiac risk did not have statistically significant correlation with the glycemic group, demonstrating that patients with higher blood glucose levels did not undergo surgeries associated with higher cardiovascular morbimortality than individuals with lower blood glucose levels.
Some studies have tried to establish the relationships among postoperative adverse events and characteristics that might be identified as predisposing factors and, therefore, creating risk indexes 3,20-24. Physical status (ASA), type of surgery, and the presence of comorbidities are frequent factors present in the constitution of those indexes.
Analysis of blood glucose levels as a function of POC and deaths is becoming more frequent among researchers around the world. Several studies using mainly patients with myocardial infarctions undergoing or not myocardial revascularization demonstrated an increase in the incidence of death and cardiovascular complications in patients with elevated blood glucose levels at admission, regardless of the presence of a history of DM 8,25-27. An association between hyperglycemia and increased mortality has also been demonstrated in studies in other populations. Krinsley 9 reported a direct correlation between mortality and progressively increasing blood glucose levels in patients both clinical and surgical, severely ill, admitted to the intensive care unit. Umpierrez et al. 28 demonstrated that hyperglycemia at the moment of admission to the hospital was an important marker of poor clinical evolution and increased mortality in hospitalized patients in a teaching hospital. In surgical patients admitted to intensive care strict plasma glucose control by implementing protocols of insulin therapy reduced significantly mortality and the incidence of postoperative complications 8,11.
Analysis of blood glucose levels as a function of the incidence of POC and deaths was the main objective of the present study. For such, it was necessary to use the concomitant analysis of parameters from the medical records that could, potentially, be associated with POC and deaths. Thus, we decided to include physical status (ASA), PSS cardiac risk, and history of DM.
The results of the present study revealed that, in patients with elevated and intermediate PSS cardiac risk, the relative risk of POC was 11.08 and 4.14 times greater, respectively, with a wide confidence interval. Patients with plasma glucose levels between 100 and 125 mg.dL-1 and greater than 126 mg.dL-1 had a probability of POC 2.11 and 3.05 times greater, respectively, regardless of the influence of other variables, with adequate confidence intervals. Such results confirmed the validity of evaluating the PSS cardiac risk as a variable associated with a greater incidence of POC 24 and indicated an elevated risk of POC even in patients with blood glucose levels between 100 and 125 mg.dL-1.
This study also demonstrated that the relative risk of deaths in patients ASA II was not significant. But the relative risk of deaths for patients ASA III/IV was 16.18 times greater, and for patients with elevated and intermediate PSS cardiac risk it was 11.69 and 6.66 times greater, respectively with wide confidence intervals. These results demonstrate the importance of evaluating physical status and the PSS cardiac risk when estimating the risk of death in this group of patients.
The results of the present study, with the limitation of having analyzed data restricted to the period of hospitalization contained in medical records, were similar to the growing concept that even glucose plasma levels below the lower limits established for the diagnosis of DM are associated with cardiovascular adverse effects 29. In fact, several studies with a high level of evidence or follow-up for more than 10 years have demonstrated the association of blood glucose levels between 100 and 125 mg.dL-1 and the increased incidence of cardiovascular morbimortality 30-32.
Identification of the glycemic profile, functional status, type of surgery, and the presence of associated comorbidities, such as DM, help guide the perioperative approach of patients, making it possible to create strategies to reduce the incidence of undesirable outcomes, such as elevated postoperative morbimortality, increased cost for health institutions, and worsening in the of quality of life of patients and emotional distress of relatives 23.
01. Dzankic KS, Pastor D, Gonzalez C et al. The prevalence and predictive value of abnormal preoperative laboratory tests in elderly surgical patients. Anesth Analg, 2001;93:301-308. [ Links ]
02. Jin F, Chung F Minimizing perioperative adverse effects in the elderly. Br J Anaesth, 2001;87:608-624. [ Links ]
03. Cook DJ, Rooke GA Priorities in perioperative geriatrics. Anesth Analg, 2003;93:1823-1836. [ Links ]
04. Brown, DR Perioperative management of the diabetic patient, em: American Society of Anesthesiologists. 56th Refresher Course Lectures and Basic Science Reviews. American Society of Anesthesiologists, 2005. CD-ROM. [ Links ]
05. Weir CJ, Murray GD, Dyker AG et al. Is hyperglycaemia an independent predictor of poor outcome after acute stroke? Results of a long-term follow up study. BMJ, 1997;314:1303-1306. [ Links ]
06. Demchuck AM, Morgenstern LB, Krieger DW et al. Serum glucose level and diabetes predict tissue plasminogen activator-related intracerebral hemorrhage in acute ischemic stroke. Stroke, 1999;30:34-39. [ Links ]
07. McCowan KC, Malhoota A, Bistrian BR Endocrine and metabolic dysfunction syndromes in the critically ill. Crit Care Clin, 2001;17:107-124. [ Links ]
08. Stranders I, Diamant M, Van Gelder RE et al. Admission blood glucose level as risk indicator of death after myocardial infarction in patients with and without diabetes mellitus. Arch Intern Med, 2004;164:982-988. [ Links ]
09. Krinsley JS Association between hyperglycemia and increased hospital mortality in a heterogeneous population of critically ill patients. Mayo Clin Proc, 2003;78:1471-1478. [ Links ]
10. Furnary AP, Zerr KJ, Grunkemeier GL et al. Continuous intravenous insulin infusions reduces the incidence of deep sternal wound infection in diabetic patients after cardiac surgical procedures. Ann Thorac Surg 2000;69:667-668. [ Links ]
11. Van den Berghe G, Wouters P, Weekers F et al. Intensive insulin therapy in critically ill patients. N Engl J Med, 2001;345: 1359-1367. [ Links ]
12. Finney SJ, Zekvaeld C, Elia A et al. Glucose control and mortality in critically ill patients. JAMA, 2003;290:2041-2047. [ Links ]
13. The Expert Commitee on the Diagnosis and Classification of Diabetes Mellitus. Follow-up report on the diagnosis of diabetes mellitus. Diabetes Care, 2003;26:3329-3330. [ Links ]
14. Almeida Filho N, Rouquayol MZ Desenhos de Pesquisa em Epidemiologia, em: Almeida Filho N, Rouquayol MZ Introdução à Epidemiologia. 3ª Ed, Rio de Janeiro, Medsi; 2002;69-214. [ Links ]
15. Word Health Organization Definition, Diagnosis and Classification of Diabetes Mellitus and its Complications. Disponível em: http://www.staff.ncl.ac.uk/philip.home/who_dmg.pdf. Acesso em: jul 2006. [ Links ]
16. Oswald GA, Smith CC, Betteridge DJ et al. Determinants and importance of stress hyperglycaemia in non-diabetic patients with myocardial infarction. BMJ, 1986;293:917-922. [ Links ]
17. Tenerz A, Lonnberg I, Berne C et al. Myocardial infarction and prevalence of diabetes mellitus: is increased casual blood glucose at admission a reliable criterion for the diagnosis of diabetes? Eur Heart J, 2001;22:1102-1110. [ Links ]
18. Bolk J, Van der Ploeg TJ, Cornel JH et al. Impaired glucose metabolism predicts mortality after a myocardial infarction. Int J Cardiol, 2001;79:207-214. [ Links ]
19. Coursin DB, Connery LE, Ketzler JT Perioperative diabetic and hyperglycemic management issues. Crit Care Med, 2004;32 (Suppl 4):116-125. [ Links ]
20. Detsky AS, Abrams HB, Forbath N et al. Cardiac assessment for patients undergoing noncardiac surgery: a multifactorial clinical risk index. Arch Intern Med, 1986;146:2131-2134. [ Links ]
21. Pedersen T, Eliasen K, Henriksen E A prospective study of mortality associated with anaesthesia and surgery: risk indicators of mortality in hospital. Acta Anaesthesiol Scand, 1990;34:176-182. [ Links ]
22. Arvidsson S, Ouchterlony J, Sjostedt L et al. Predicting postoperative adverse events: clinical efficiency of four general classification systems the project perioperative risk. Acta Anaesthesiol Scand 1996;40:783-91. [ Links ]
23. Liu L, Leung J Predicting adverse postoperative outcomes in patients aged 80 years or older. J Am Geriatr Soc, 2000;48:405-412. [ Links ]
24. Eagle KA, Berger PB, Calkins H et al. ACC/AHA guideline update for perioperative cardiovascular evaluation for noncardiac surgery executive summary: A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee to Update the 1996 Guidelines on Perioperative Cardiovascular Evaluation for Noncardiac Surgery). Anesth Analg, 2002;94:1052-1064. [ Links ]
25. Norhammar AM, Rydén L, Malmberg K Admission plasma glucose: independent risk factor for long term prognosis after myocardial infarction even in nondiabetic patients. Diabetes Care, 1999;22:1827-1831. [ Links ]
26. Capes SE, Hunt D, Malmberg K et al. Stress hyperglycaemia and increased risk of death after myocardial infarction in patients with and without diabetes: a systematic overview. Lancet North Am Ed, 2000;355:773-778. [ Links ]
27. Norhammar A, Tenerz A, Nilsson G et al. Glucose metabolism in patients with acute myocardial infarction and no previous diagnosis of diabetes mellitus: a prospective study. Lancet 2002; 359:2140-2145. [ Links ]
28. Umpierrez GE, Isaacs SD, Bazargan N et al. Hyperglycemia: an independent marker of in-hospital mortality in patient with undiagnosed diabetes. J Clin Endocrinol Metab, 2002;87:978-982. [ Links ]
29. Gerstein H Is glucose a continuous risk factor for cardiovascular mortality? Diabetes Care, 1999;22:659-660. [ Links ]
30. Fuller JH, Shipley MJ, Rose G et al. Coronary-heart-disease risk and impaired glucose tolerance: the Whitehall study. Lancet, 1980;8183:1373-1376. [ Links ]
31. Bjornholt JV, Erikssen G, Aaser E et al. Fasting blood glucose: an underestimated risk factor for cardiovascular death: results from a 22-year follow up of healthy non-diabetic men. Diabetes Care, 1999;22:45-49. [ Links ]
32. Coutinho M, Gerstein HC, Wang Y et al. The relationship between glucose and incident cardiovascular events: a metaregression analysis of published data from 20 studies of 95,783 individuals followed for 12.4 years. Diabetes Care, 1999;22: 233-240. [ Links ]
Dra. Lígia Andrade da Silva Telles Mathias
Alameda Campinas, 139/41
01404-000 São Paulo, SP
Submitted em 30
de setembro de 2006
Accepted para publicação em 20 de agosto de 2007
* Received from CET/SBA, Serviço de Anestesiologia da Irmandade da Santa Casa de Misericórdia de São Paulo (ISCMSP), São Paulo, SP