BJA Advance Access originally published online on October 24, 2007
British Journal of Anaesthesia 2007 99(6):824-829; doi:10.1093/bja/aem307
Determinants of outcome in critically ill octogenarians after surgery: an observational study
1 Department of Anaesthesia, Royal Devon and Exeter Hospital, Barrack Road, Exeter EX2 5DW, UK
2 Department of Anaesthesia, Frenchay Hospital, Frenchay Park Road, Bristol BS16 1LE, UK
3 Department of Anaesthesia, Royal United Hospital, Combe Park, Bath BA1 3NG, UK
4 Bristol NHS Research and Development Support Unit, The Coach House, Southmead Hospital, Westbury on Trym, Bristol BS 10 5NB, UK
* Corresponding author. E-mail: fordpeter{at}doctors.org.net
Accepted for publication August 17, 2007.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Background: The population in the UK is growing older and the number of elderly patients cared for on intensive care units (ICU) is increasing. This study was designed to identify risk factors for mortality in critically ill patients of >80 yr of age after surgery.
Methods: We identified 275 patients, aged 80 yr or greater, admitted to the ICU after surgery. After exclusions, 255 were selected for further analysis. Multivariate analysis was then performed to determine the covariates associated with hospital mortality.
Results: The overall ICU and hospital mortality was 20.4% and 33.3%, respectively. Patients who received i.v. vasoactive drugs on days 1 and 2 had hospital mortality of 54.4% and 60.5%, respectively. Multivariate analysis showed that requirement for i.v. vasoactive drugs within the first 24 h on ICU [odds ratio (OR) 4.29; 95% CI, 2.35–7.84, P<0.001] and requirement for i.v. vasoactive drugs for a further 24 h (OR 3.63; 95% CI, 1.58–8.37, P<0.01) were associated with hospital mortality. The requirement for i.v. vasoactive drugs was also strongly associated with hospital mortality in all the subgroups studied (elective surgery, emergency surgery, and emergency laparotomy).
Conclusions: For patients aged 80 yr and more, admitted to ICU after surgery, the requirement for i.v. vasoactive drugs in the first and second 24 h was the strongest predictor of hospital mortality.
Keywords: age factors; complications, death; intensive care; surgery, postoperative period
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The population in the UK is growing older. In 1951, those aged 85 yr and more constituted only 1.6% of the population, in 2003 this had risen to 5.5%, and by 2031 it is predicted to grow to 7.9%.1 This ageing population is consuming increasing amounts of healthcare resources, with the number of elderly patients admitted to intensive care (ICU) growing progressively.2 In the USA, more than half of all intensive care days are incurred by patients older than 65 yr of age.3 Recent attempts at healthcare cost containment have fuelled discussions on rationing, using age as a discriminator.4 Whether or not rationing based on age becomes explicit, age is already having a negative impact on the decision to admit to ICU.5–7 Justification for this is based on the physiological decline and the increase in co-morbidities seen with age. The incorrect assumption from these observations is that the chances of survival from a serious illness in the elderly are unlikely: outcome studies examining this suggest that age is not a good predictor of outcome on ICU.8–11 In this study, we sought to identify other factors taken from the first 48 h on ICU which would predict length of stay and mortality in the very elderly.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
We studied a retrospective, unmatched cohort of patients admitted to the ICU of the Royal United Hospital, Bath, UK between April 1, 1998 and March 2003. The Royal United Hospital serves a population of 550 000 people and has an 11-bedded ICU. The ICU admits levels 2 and 3 patients (level 2 patients require close observation or support for a single failing organ system and include those stepping down from level 3; level 3 patients require advanced respiratory support or basic respiratory support plus support for two other organ systems).12 All patients admitted to the ICU had their variables entered into a clinical database from which data were incorporated into the national case mix programme database coordinated by the Intensive Care National Audit and Research Centre. The target population for this study was patients more than 80 yr of age admitted to the ICU after a surgical procedure. Written consent and formal ethics committee approval were required as suggested by the chair of the committee.
We identified all patients aged 80 yr or more, who had been admitted to the ICU after a recent surgical procedure. The following data were then recorded for each patient: age, gender, nature of surgical procedure, urgency of surgery, Acute Physiology and Chronic Health Evaluation 2 (APACHE II) score, requirement for ventilation or i.v. vasoactive drugs in the first 24 h, and in the second 24 h on the ICU, survival status at ICU, and hospital discharge. The database was carefully reviewed and, where necessary, supplemental data were extracted from an operating theatre database and patients' paper records. For this study, patients undergoing elective or scheduled operations, as defined by the National Confidential Enquiry into Perioperative Deaths, were categorized as elective and those undergoing urgent or emergency operations were categorized as emergency. Patients were recorded as ‘receiving i.v. vasoactive drugs in their first and second 24 h period’ if they received any vasoactive drug [defined as inotropes (e.g. dobutamine) or vasopressors (e.g. noradrenaline)] by infusion at any point during their first or second 24 h, respectively. Patients were recorded as ‘receiving ventilation in their first and second 24 h period’, if they received any intermittent positive pressure ventilation (including mandatory or assisted modes) via a tracheal tube or trachoestomy at any point during their first or second 24 h, respectively. Patients readmitted to the ICU had only their first admission data used for the APACHE II score and their admissions amalgamated for length of stay and survival. Patients who stayed on the ICU for less than 0.2 days and who did not generate an APACHE II score were not included.
The primary outcome of the study was hospital mortality; secondary outcomes included ICU mortality and lengths of stay for hospital and ICU. ICU mortality was defined as survival of a patient at ultimate discharge from the ICU and hospital mortality was defined as survival at discharge or transfer from our hospital.
All data were analysed using Stata Statistical software (version 9.1, ©Statacorp). Continuous data were summarized using mean (SD) and medians (IQR). Univariable analysis using logistic regression was used to identify statistically significant prognostic covariates associated with hospital mortality and ICU length of stay (length of stay was divided above and below 3 days and the analysis conducted only on ICU survivors). The crude odds ratios and 95% confidence intervals were adjusted for potential confounding variables: gender, age, and severity of illness (measured using the APACHE II score). A further multivariable model was used to identify independent predictors of hospital mortality. All covariates found to be significant in the univariable analysis were included except APACHE II score. For each subgroup, multivariable analysis was performed twice, for the first 24 h of an admission (i.e. all patients) and secondly for those surviving into the second 24 h of an ICU admission. For length of stay, covariates with an odds ratio <1 imply a trend towards a stay <3 days on the ICU. Conversely for hospital mortality, covariates with an odds ratio >1 imply an increased chance of dying. For both univariable and multivariable analysis, differences were considered significant at P<0.05.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
During the study period, 275 patients more than 80 yr of age were identified as being admitted to the ICU after surgery. This represented 8.8% of all admissions during the 5 yr period. Eleven patients were excluded for being on the ICU for less than 0.2 days and did not generate an APACHE II score and nine admissions were recognized as being readmissions. Therefore, 255 patients were submitted for further analysis. The characteristics of the study population are summarized in Table 1. The median age was 83.0 yr (IQR 81–86), 5.8% were more than 90 yr of age. Sixty per cent of admissions followed an emergency operation. There were 137 (53.7%) laparotomies, 54 (21.2%) abdominal aortic aneurysm repairs, 8 (3.2%) other vascular cases, 25 (9.8%) orthopaedic cases, and 31 (12.1%) other cases. As shown in Table 1, the overall ICU and hospital mortality was 20.4% and 33.3%, respectively. For patients who received i.v. vasoactive drugs on day 1 of ICU admission, their ICU mortality was 39.8% and hospital mortality was 54.4%. For patients receiving i.v. vasoactive drugs on day 2 of their ICU admission, their ICU mortality was 44.7% and their hospital mortality was 60.5%.
|
The unadjusted, univariate analysis revealed the following significant covariates: urgency of surgery, type of operation, APACHE II score, requirement for ventilation on day 1, use of ventilation on day 2 of an ICU admission, and use of vasoactive drugs on days 1 and 2 of an ICU admission. After adjusting for age, sex, and severity of illness the following remained significant: urgency of surgery [adjusted OR (adjOR) 2.79; 95% CI, 1.49–5.2; P=0.001], operation type (e.g. AAA, adjOR 0.31; 95% CI, 0.13–0.70; P=0.001), use of i.v. vasoactive drugs on day 1 of an ICU admission (adjOR 3.81; 95% CI, 2.11–6.88; P<0.001), use of i.v. vasoactive drugs on day 2 of an ICU admission (adjOR 5.81; 95% CI, 3.1–10.89; P<0.001), and use of ventilation on day 2 of an ICU admission (adjOR 3.73; 95% CI, 1.95–7.12; P<0.001).
For those surviving ICU (n=203), a univariate analysis was performed against length of stay on ICU, divided into above or below 3 days. Significant predictors of a stay on ICU for more than 3 days (after adjustment for age, sex, and severity of illness) included urgency of surgery (adjOR 0.42; 95% CI, 0.21–0.83; P=0.01), use of ventilation on day 1 of an ICU admission (adjOR 0.41; 95% CI, 0.23–0.83; P=0.01), use of ventilation on day 2 of an ICU admission (adjOR 0.18; 95% CI, 0.08–0.41; P<0.001), use of i.v. vasoactive drugs on day 1 of an ICU admission (adjOR) 0.13; 95% CI, 0.06–0.27; P<0.001), use of i.v. vasoactive drugs on day 2 of an ICU admission (adjOR 0.1; 95% CI, 0.04–0.22; P<0.001).
A multivariate analysis using logistic regression was used to evaluate the independent role of each covariate in hospital mortality. Multivariate analysis was conducted on all patients included in the study followed by a further three subgroups: after elective surgery, emergency surgery, and emergency laparotomy. All covariates included in the univariate analysis were submitted to multivariate analysis except APACHE II score because of co-linearity with some of the other covariates. The results of the multivariate analysis are shown in Tables 2–5.
|
|
|
|
For all the patients in the study, those who received i.v. vasoactive drugs within the first 24 h on ICU had a 4.3-fold increased mortality (OR 4.29; 95% CI, 2.35–7.84; P<0.001). Patients who received vasoactive drugs on day 2 had a 3.6-fold increase in mortality (OR 3.63; 95% CI, 1.58–8.37; P<0.01). Patients who received ventilation extending into the second 24 h on ICU had a two-fold increase in mortality (OR 2.1; 95% CI, 0.94–4.7), which was just outside our limit of statistical significance (P=0.07).
After an elective operation, patients who received i.v. vasoactive drugs within the first 24 h on ICU had a 5.8-fold increase rate of death (OR 5.8; 95% CI, 1.86–18.10; P=0.002); for those who received ventilation on the second day of ICU mortality increased 6.9-fold (OR 6.88; 95% CI, 1.5–45.06; P=0.04).
After an emergency operation, patients who received i.v. vasoactive drugs within the first 24 h on ICU had a four-fold increased mortality rate (OR 4.06; 95% CI, 1.93–8.53; P<0.001). Those who received vasoactive drugs on the second day of ICU had a four-fold increase in mortality (OR 3.96; 95% CI, 1.53–10.21; P=0.004).
Finally, after an emergency laparotomy, patients who received i.v. vasoactive drugs within the first 24 h on ICU were 3.9 times more likely to die (OR 3.85; 95% CI, 1.64–9.02; P=0.002). No significant prognostic factors were determined by the model on day 2.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This study has shown a consistent association between the requirement of i.v. vasoactive drugs in the first 48 h and mortality. The use of i.v. vasoactive drugs in the first two 24 h periods after admission was associated with a 3.6 to 5.8-fold increase in the likelihood of dying, depending on the urgency of the surgery. Ventilation extending into the second 24 h was generally associated with an increase in mortality and after elective operations, a 6.9-fold increase in the risk of death. Notably, the use of ventilation in the first 24 h after admission was not associated with increased mortality. Length of stay beyond 3 days was associated with urgency of surgery, use of i.v. vasoactive drugs, and ventilation in the first two 24 h periods. Nowhere in these analyses did we find an association between increasing age and mortality. This confirms the previous studies on the critically ill elderly that age per se is not a strong predictor of death on ICU.
Over the last decade, there has been increasing numbers of studies focusing on the outcome of the elderly admitted to ICU. Mortality rates have been found to be high. However, many studies have shown that the severity of illness rather than age itself is the strongest predictor of death. Chelluri and colleagues made a comparison between 54 patients aged 75 yr and older and 43 patients aged between 65 and 74 yr admitted to a mixed ICU. Long-term survival was similar between the two age groups but differed significantly between those with low and those with high APACHE II scores.8 Another similar study by Wu and colleagues compared 130 patients aged 75 yr and older and those aged between 55 and 65 yr, admitted over a 30 month period to a medical ICU. Using logistic regression, the risk of dying was associated with a modified APACHE II score; increased age did not predict mortality.9 A longitudinal study conducted by Kass and colleagues followed patients over the age of 85 yr admitted to a mixed ICU. In that study, age was not associated with mortality, instead severity of illness determined by the number of organs failing was significantly associated with outcome at: ICU discharge, 30 day post-hospital discharge, and 1 yr. In patients with more than two organ system failure, there was 100% mortality at 1 yr.10 Severity of illness predicting death more accurately than age applies also to subgroups of patients on ICU. Shabot and colleagues evaluated 1039 trauma patients admitted to a surgical ICU, of which 9.5% were more than 65 yr of age. Age was a poor determinant of survival whereas Injury Severity Score (ISS), Trauma Score (TS), and the Simplified Acute Physiology Score (SAPS) were good determinants of survival.11
Few studies have concentrated specifically on the outcomes after surgery. Stephan and colleagues investigated the outcome of patients on a surgical ICU. They included 106 patients more than the age of 75 and found an ICU mortality of 31.0% and a hospital mortality of 42.4%.13 Using a multivariate model, hospital mortality was associated with an ICU-acquired nosocomial infection, mechanical ventilation for more than 24 h, modified APACHE II score, duration of preoperative hospitalization, and underlying disease. Similarly, 140 patients more than the age of 90 were studied by Margulies and colleagues.14 ICU mortality was found to be 4.3% and hospital mortality 17.1%. Udekwas and colleagues evaluated 672 patients aged 70 yr or older admitted to a surgical ICU. The study included 65.5% of elective cases and 50.7% of vascular surgical patients. Hospital mortality was 12.1% and was found to be associated with increasing age, severity of illness, and length of stay.15 In patients older than 85 yr of age, Yosylius and colleagues found an ICU and hospital mortality of 29% and 45%, respectively. Although this was a mixed ICU population, 73% were admitted after an operation. Independent factors associated with mortality were: impaired level of consciousness, infection on admission, ICU-acquired infection, and severity of illness.16 The gross variations in survival between the studies are explained by their heterogeneity. The studies differed in age criteria, severity of illness, types of operations, and country of origin.
Our study has limitations. First, it is an observational study and therefore there is potential error from unmeasured confounding factors. Secondly, it is assumed in our study that patients who received vasoactive drugs or ventilation ‘required it’. As this study is retrospective, the study can have had no influence on decisions to use these interventions and it is likely that the decision to use or not use these interventions is similar in both groups. Thirdly, our subgroup analysis is inevitably of smaller groups and fewer variables appear significantly associated with poorer outcome. Larger studies of these subgroups may be indicated. Lastly, the sickest patients may be subject to a self-fulfilling prophecy in that treatment is withdrawn before the full duration of the disease process. We did not measure the number of patients who had their treatment withdrawn before death. However, our results suggest that increased intervention (rather than withdrawal of such treatment) on days 1 and 2 after admission is associated with increased length of stay and mortality. The strength of the study lies in the external validation. The Royal United Hospital is a large general hospital and performs general surgical, orthopaedic, vascular, urological, and gynaecological operations. Similar cohorts of patients are admitted to many other hospitals across the UK.
In conclusion, this study found that for patients aged 80 and more, admitted to ICU after surgery the strongest association with mortality was the use of i.v. vasoactive drugs during both of their first 2 days on ICU. Length of stay on ICU increased after emergency surgery and in patients who received i.v. vasoactive drugs and ventilation on both the first and second days of admission. After elective operations, ventilation beyond the first 24 h was also a significant predictor of death.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1 www.statistics.gov.uk. Accessed March 2007.
2 Jakob SM, Rothen HU. Intensive care 1980–1995: change in patient characteristics, nursing workload and outcome. Intensive Care Med (1997) 23:1165–70.[CrossRef][Web of Science][Medline]
3 Angus DC, Kelley MA, Schmitz RJ, White A, Popovich J. Current and projected workforce requirements for care of the critically ill and patients with pulmonary disease: can we meet the requirements of an aging population? JAMA (2000) 284:2762–70.
4 Levinsky NG. Age as a criterion for rationing health care. N Engl J Med (1990) 322:1813–5.[Web of Science][Medline]
5 Nuckton TJ, List ND. Age as a factor in critical care unit admissions. Arch Intern Med (1995) 155:1087–92.
6 Garrouste-Orgeas M, Montuclard L, Timsit J-F, Misset B, Christias M, Carlet J. Triaging patients to the ICU: a pilot study of factors influencing admission decisions and patient outcomes. Intensive Care Med (2003) 29:774–81.[Web of Science][Medline]
7 Hamel MB, Teno JM, Goldman L, et al. Patient age and decisions to withhold life sustaining treatments from seriously ill, hospitalized adults. Ann Intern Med (1999) 130:116–25.
8 Chelluri L, Pinsky MR, Donahoe MP, Grenvik A. Long-term outcome of critically elderly patients requiring intensive care. JAMA (1993) 269:3119–23.
9 Wu AW, Rubin HR, Rosen MJ. Are elderly people less responsive to intensive care? JAGS (1990) 38:621–7.
10 Kass JE, Castriotta RJ, Malakoff F. Intensive care unit outcome in the very elderly. Crit Care Med (1992) 20:1666–71.[Web of Science][Medline]
11 Shabot MM, Johnson CL. Outcome from critical care in the oldest old trauma patients. J Trauma (1995) 39:254–60.[Web of Science][Medline]
12 Level of care document. Accessed March 2007.
13 Stephan F, Cheffi A, Bonnet F. Nosocomial infections and outcome of critically ill elderly patients after surgery. Anesthesiology (2001) 94:407–14.[CrossRef][Web of Science][Medline]
14 Margulies DR, Lekawa ME, Bjerke S, Hiatt JR, Shabot MM. Surgical intensive care in the nonagenarian. Arch Surg (1993) 128:753–8.
15 Udekwu P, Gurkin B, Oller D, Lapio L, Bourbina J. Quality of life and functional level in elderly patients surviving surgical intensive care. J Am Coll Surg (2001) 193:245–9.[CrossRef][Web of Science][Medline]
16 Yosylius S, Sipylaite J, Ivaskevicius J. Determinants of outcome in elderly patients admitted to the intensive care unit. Age Ageing (2005) 34:157–62.
Read all E-letters
CiteULike 
Connotea 
Del.icio.us What's this?
E-letters:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||