Free
Education  |   March 2001
Nosocomial Infections and Outcome of Critically Ill Elderly Patients after Surgery
Author Affiliations & Notes
  • François Stéphan, M.D., Ph.D.
    *
  • Ali Cheffi, M.D.
  • Francis Bonnet, M.D.
  • *Assistant Professor, †Resident, ‡Professor and Chairman.
  • Received from the Service d’Anesthésie-Réanimation Chirurgicale, Hôpital Tenon, Paris, France.
Article Information
Education
Education   |   March 2001
Nosocomial Infections and Outcome of Critically Ill Elderly Patients after Surgery
Anesthesiology 3 2001, Vol.94, 407-414. doi:
Anesthesiology 3 2001, Vol.94, 407-414. doi:
ArticlePlus
Click on the links below to access all the ArticlePlus for this article.
Please note that ArticlePlus files may launch a viewer application outside of your web browser.
THE elderly population is increasing in terms of both absolute numbers and proportion of the overall population, and, in the future, this population will use a greater proportion of health care resources. Elderly patients currently constitute 12.5 1 to 58%2 of patients in the intensive care unit (ICU). The ageing process leads to variable decline of physiologic and morphologic functions and differential changes in other organ systems, which should intuitively be expected to lead independently to increased morbidity, greater resource consumption, and a higher mortality rate for the elderly critically ill patients. 3 Many investigators have studied the effect of age on outcome of critical illness with variable results. Although the effect of age on outcome is unclear, severity of illness has been shown to play a significant and consistent role in determining the outcome. 4–6 It is difficult to compare these studies because the age of elderly patients has varied from more than 60 to more than 75 yr and older, and heterogeneous groups of patients were studied, in which patients admitted to ICU after surgery constituted only a small proportion. Moreover, the highest rates of nosocomial infections are observed in the ICU, which is where the most severely ill patients are treated and where the highest mortality rates are observed. The adverse effects of nosocomial infections have previously been estimated in terms of mortality, morbidity, and other consequences, such as economic impact. 7 Nonetheless, the relation between older age and nosocomial infections is difficult to evaluate because of a shortage of previous studies. However, it has been postulated that elderly hospitalized patients are more likely to experience a nosocomial infection. 3,8,9 In view of the consequences of ICU-acquired nosocomial infections, it is therefore important to try to answer this straightforward problem. The aims of this study were therefore: to determine whether elderly surgical patients constitute a population at increased risk of nosocomial infection and to assess the influence of age on outcome in this specific group of patients.
Patients and Methods
Study Population
The institutional clinical investigation committee (Comité d’Ethique, Hôpital Tenon, Paris, France) approved the study protocol. No informed consent was mandatory, because this observational study did not modify current diagnostic or therapeutic strategies.
The study was conducted in an eight-bed surgical ICU receiving patients mainly from the operating room, recovery room, and emergency room. During the study period, from November 1, 1995 through December 31, 1997, 646 patients were admitted to the unit, including 406 postoperative patients who constituted the study population. Our institution has no predefined age or specific diagnostic criteria for referral or admission to the ICU.
In the absence of a clearly defined threshold for “elderly” in the medical literature, a cut-off point of more than 75 yr was chosen as the most widely accepted in previous studies. 1,10–13 We also divided the other patients into two groups: young patients (< 60 yr) and intermediate age patients (60–75 yr).
Data Collection
Data were collected prospectively for each patient. Severity of illness was measured by the Acute Physiology and Chronic Health Evaluation II (APACHE II) score, 14 modified APACHE II score (which did not include age), 11,13 and the Organ Dysfunction and Infection model. 15 Various other variables were noted: age; sex; type of surgery; main diagnostic categories leading to ICU admission; severity of underlying medical conditions at admission, stratified according to the criteria of McCabe and Jackson as likely to be fatal, ultimately fatal, and nonfatal, 16 and preadmission health status 17 (obtained directly from the patient whenever possible and from relatives when necessary); presence or absence of sepsis as previously described 18; and need and duration of mechanical ventilation. Chronic obstructive pulmonary disease was diagnosed according to the standards of the American Thoracic Society. 19 Obesity was defined as a body mass index of more than 30 kg/m2for men and more than 28.6 kg/m2for women. A history of a recent (within 3 months) loss of 10% or more of body weight signified substantial protein and calories malnutrition. Therapeutic activity was evaluated using the Omega score. 20 The Omega score is composed of therapeutic items scored from 1 to 10 points and divided into three categories as follows: category 1, items entered only at the time of their first application; category 2, items entered at each application; and category 3, items entered every day of application. The total score, which covers the entire length of stay, is calculated by adding the points obtained in the three categories (see 1).
ICU-acquired Infections
Information concerning ICU-acquired infections was also collected on the survey form (type, microbiologic data, and date of occurrence). ICU-acquired infection was defined as an infection that began at least 48 h after ICU admission. In patients assisted by mechanical ventilation, the diagnosis of nosocomial pneumonia was considered when a new and persistent lung infiltrate developed and the patient had purulent tracheal secretions, confirmed by a bacterial culture of the protected specimen brush of more than 103colony-forming units (CFU)/ml, bronchoalveolar lavage of more than 104CFU/ml, positivity of bronchoalveolar lavage on direct examination, defined as more than 5% of cells containing intracellular bacteria, or a combination thereof. 21 In patients breathing spontaneously, the diagnosis was considered when they had a compatible chest radiograph and purulent sputum, with Gram stain and sputum culture documenting the presence of microorganisms. Diagnosis of central venous catheter-related infection was confirmed by a positive quantitative tip culture with a significant threshold of 103CFU/ml. 22 Diagnosis of primary bacteremia was confirmed by at least one positive blood culture (two or more blood cultures when coagulase-negative staphylococci were isolated) without another site simultaneously infected with the same organism. A urinary tract infection was defined by the combination of the following two criteria: pyuria (≥10 leukocytes/mm3) and a urine culture growing 105CFU/ml in patients with clinical signs of infection (fever > 38°C, leukocytosis, abnormal macroscopic appearance of urine, and presence of urinary nitrites). Sinusitis was suspected in patients with fever, purulent nasal secretions, or both, who had a radiologic opacification of the maxillary sinuses, and was confirmed by a sinus aspirate containing more than five altered polymorphonuclear leukocytes per oil-immersion field and by a positive microbiologic culture with a quantitative threshold of 104CFU/ml. 23 The diagnosis of surgical wound infections was based on clinical examination and confirmed by microbiologic analysis of specimens.
Resistant bacteria were defined as ticarcillin-resistant Pseudomonas aeruginosa  , Acinetobacter baumannii  , and Stenotrophomonas maltophilia  ; extended-spectrum lactamase-producing Enterobacteriaceae; and methicillin-resistant Staphylococcus aureus  .
Treatment of infected patients did not differ between elderly and younger patients. A broad-spectrum β-lactam antibiotic plus aminoglycoside with or without vancomycin were the first-line agents for empiric treatment. Fluconazole was the first-line agent in case of Candida  infection. Antibiotics with a narrower spectrum of activity were systematically used, based on identification and susceptibility test results for pathogens cultured. The recommended duration of therapy was 14 days for ventilator-associated pneumonia and bloodstream infections; 14 days for peritonitis and intraabdominal abscesses; 6 weeks for mediastinitis, prosthetic vascular or joint arthroplasty infections; 14 or 21 days for pyelonephritis; and 7 days for lower urinary tract infection. When decided, second surgery after initial admission was scheduled within 12 h maximum after diagnosis.
The incidence rate was defined as the number of new cases of infection divided by the number of patients studied. The incidence density was defined as the number of new cases of infection divided by the total number of patient days in the study population. The device-associated incidence density was defined as the number of new cases of site infection divided by the number of device-exposed days in the study population.
Outcome Measures
Length of stay and mortality in the ICU and in the hospital was recorded. Place of residence after hospital discharge was also recorded.
Statistical Analysis
Data were computerized and analyzed using Statview 5.0 statistical packages (SAS Institute Inc., Cary, NC). We expressed continuous variables as the mean (± standard deviation) or as the median and 25th–75th percentiles if their distribution was skewed. Chi-square tests were used to compare proportions and rates. Comparisons between the three groups were performed by analysis of variance followed by a Scheffé F test for post hoc  comparisons of quantitative variables. The Kruskal-Wallis test was used to analyze continuous variables with a nonnormal distribution. Multivariate modelling in the form of Cox proportional hazard analysis was used to evaluate potential risk factors for ICU-acquired nosocomial infection or for ICU or hospital death. Variables associated with ICU-acquired nosocomial infections or ICU or hospital death in the univariate analysis with a P  value less than 0.05 were entered into the model. The modified APACHE II scores were used because proportional hazards models already assessed the effect of age. Hazard ratios (HR) and 95% confidence intervals (95% CI) were calculated for all significant predictors. Statistical significance was defined as a P  value of 0.05 or less.
Results
Demographic Characteristics
Four hundred six patients admitted to ICU after surgery were prospectively evaluated. As shown in table 1, the proportion of men was lower among elderly patients. Elderly patients also had more severe illnesses, as documented by higher APACHE II scores, with or without age adjustment (P  < 0.0001). Surgical procedures and the respective percentages of the 106 elderly patients who underwent surgery can be found in a table on the Anesthesiology Web site. The reasons for ICU admission were diverse (table can be found on the Anesthesiology Web site) and were distributed unevenly between the three age groups (P  < 0.005).
Table 1. Characteristics of the Study Population
Image not available
Table 1. Characteristics of the Study Population
×
Nosocomial Infections
One hundred three of the 406 patients experienced at least one nosocomial infection, corresponding to a global incidence rate of 25.0 infections per 100 admissions (table 2). Among the 103 patients with nosocomial infection, 42 (41.0%) had a single infection, 30 (29.0%) had two infections, and 31 (30.0%) had three or more infections. No differences were noted between the three groups regarding the percentage of multiple infections.
Table 2. Nosocomial Infections and Selected Risk Factors*
Image not available
Table 2. Nosocomial Infections and Selected Risk Factors*
×
Nosocomial infections are described in table 2. Although differences were observed between the three groups, except for bloodstream infections, no statistically significant difference was found regarding the rate of nosocomial infection between young patients and elderly patients. Nosocomial pneumonia was the most frequent infection, closely followed by infections of surgical site, which included secondary peritonitis (n = 12), intraabdominal abscesses (n = 9), pulmonary empyema (n = 6), mediastinitis (n = 5), prosthetic vascular infections (n = 4), hip replacement infections (n = 4), and renal catheterization infections (n = 2).
A total of 229 microorganisms were isolated, with 30.0% of Gram-positive cocci, 60.0% of Gram-negative bacilli, and 9.0% of Candida  species (table can be found on the Anesthesiology Web site). Nine infections (22.0%) were polymicrobial in the elderly patients, 21 (21.0%) in intermediate age patients, and nine (22.0%) in young patients (not significant). No significant differences in the frequency distribution of the various microorganisms isolated or emergence of resistant bacteria were observed between the three groups.
To evaluate whether age of more than 75 yr contributes to ICU-acquired nosocomial infection, a multivariate Cox model was built, taking into account variables univariately associated with ICU-acquired nosocomial infection. Independent risk factors were malnutrition (HR, 1.96; 95% CI, 1.04–3.70), presence of a central venous catheter (HR, 1.72; 95% CI, 1.27–2.34), unscheduled surgery (HR, 1.30; 95% CI, 1.0–1.70), length of mechanical ventilation (1-day increments; HR, 1.25; 95% CI, 1.19–1.32), Organ Dysfunction and Infection scoring system (1-point increment; HR, 1.17; 95% CI, 1.0–1.37), and modified APACHE II (1-point increments; HR, 1.04; 95% CI, 1.01–1.06). However, age more than 75 yr was not identified as an independent predictor (HR, 0.90; 95% CI, 0.67–1.21).
Duration of Mechanical Ventilation, Therapeutic Activity, Intensive Care Unit Stay, and Hospital Stay
A higher percentage of elderly and intermediate age patients required mechanical ventilation compared with young patients (table 2). Duration of mechanical ventilation was similar between young and elderly patients: The median time spent on the ventilator was 1 day (range, 0–2 days) for young patients and 1 day (range, 1–5 days) for elderly patients (not significant). However, the duration of mechanical ventilation was longer in intermediate age patients than in the other two groups (2 days; range, 1–13). Six elderly patients (6.0%), 14 intermediate age patients (10.0%), and 10 young patients (6.0%) underwent tracheostomy (not significant).
The median length of stay in the ICU was comparable between young and elderly patients, even after exclusion of patients with ICU death (table 3). Omega scores were also statistically different between the three age groups (table 3).
Table 3. Outcome Variables and Therapeutic Activity*
Image not available
Table 3. Outcome Variables and Therapeutic Activity*
×
In-hospital Mortality and Outcome
Intensive care unit and total in-hospital mortality rates are summarized in table 3. After ICU discharge, 12 additional elderly patients died (16%), compared with 11 intermediate age patients (11%) and two young patients (P  = 0.005).
To evaluate whether age more than 75 yr contributes to ICU mortality, a multivariate Cox model was built, taking into account variables univariately associated with ICU mortality. Independent risk factors were ICU-acquired nosocomial infection (HR, 4.0; 95% CI, 2.7–5.87), mechanical ventilation more that 24 h (HR, 2.43; 95% CI, 1.76–3.37), existence of sepsis (HR, 2.18; 95% CI, 1.53–3.10), American Society of Anesthesiologists (ASA) score more than II (HR, 1.41; 95% CI, 1.01–1.97), modified APACHE II score (1-point increments; HR, 1.03; 95% CI, 1.01–1.05). However, age more than 75 yr was not identified as an independent risk factor of death in ICU (HR, 0.95; 95% CI, 0.71–1.28).
Likewise, to evaluate whether age more than 75 yr contributes to hospital mortality, a multivariate Cox model was built, taking into account variables univariately associated with hospital mortality. Independent risk factors were ICU-acquired nosocomial infection (HR, 1.8; 95% CI, 1.3–2.7), mechanical ventilation more than 24 hr (HR, 1.72; 95% CI, 1.22–2.42), modified APACHE II score (1-point increments; HR, 1.03; 95% CI, 1.0–1.04), duration of preoperative hospitalization (1-day increments; HR, 1.03; 95% CI, 1.01–1.05), underlying disease ultimately or rapidly fatal (HR, 1.48; 95% CI, 1.22–2.43). Orthopedic surgery was a protective factor (HR, 0.45; 95% CI, 0.29–0.71). However, age more than 75 yr was not identified as an independent risk factor of death in hospital (HR, 1.28; 95% CI, 0.94–1.74).
As shown in table 3, place of residence at hospital discharge varied considerably between elderly and younger patients. Only 24% of elderly patients were able to return home, and nearly three quarters of elderly patients had to be treated in medium- or long-term care hospitals.
Discussion
Our results suggest, first, that elderly surgical patients admitted to the ICU are not a population at increased risk of nosocomial infection contrary to patients more than 60 yr old; second, despite their higher mortality rates in ICU and in hospital, age more than 75 yr does not appear to be a determinant risk factor of mortality after controlling for other risk factors.
It is very difficult to define elderly patients, because age is an insensitive index. The cut-off point was selected after studying demographic and epidemiologic data, and this cut-off point is one of the most frequently used in recent studies, 1,10–13 although some authors had chosen a lower cut-off point of 65–75 yr. 6,24–26 However, because the lack of effect of age more than 75 yr observed in this study could be strictly an artefact of definition, we stratified the data into three groups: less than 60 years; between 60 and 75 years; and more than 75 years.
Nosocomial infections are a major cause of morbidity and mortality in ICU patients. Many factors influence the risk of nosocomial infection in these patients, including underlying diseases, severity of illness, type of ICU, duration of ICU stay, and number, type, and duration of invasive devices and procedures. 7,27,28 Nosocomial infection rates are highest in the surgical ICU patients, and are approximately 35%. 27 The distribution of types of infection is in the range of previous reports in medical or surgical ICUs, despite some differences concerning the definition used. 27–29 Likewise, microbiologic data are in accordance with previous studies, 27–29 with aerobic Gram-negative bacilli accounting for most isolates. S. aureus  accounted for only 11%, which is close to the frequency in surgical ICUs. 27 
Increasing age is associated with increasing risk of acquiring a nosocomial infection. 3,9 In an acute-care teaching hospital, Saviteer et al.  8 demonstrated a steadily increasing nosocomial infection rate with increasing age, and another study 24 suggested that nosocomial infections occur more frequently among elderly than younger patients. However, this association lost statistical significance after controlling for potential confounding factors. 30 Older age was identified as an independent risk factor of nosocomial infections, with an odds ratio equal to 1.54 in a prospective survey in five French ICUs, 29 but not in three other studies. 27,28,31 Results focusing on ventilator-associated pneumonia are also conflicting. Whereas some studies identified older age as a potential risk factor, 32,33 others failed to demonstrate such a relation. 34–36 The possible reasons for discordant results in these studies are myriad and include enrolment of various patient populations, inclusion of different risk factors in the analysis, variable definitions of ICU-acquired infection, and different analytic methods.
Our study and other studies 7,28 have shown that the severity of underlying disease measured within the first 24 hr of intensive care is a risk factor for nosocomial infection. In the European Prevalence of Infection in Intensive Care study, beyond an APACHE II score of 15, the percentage of patients with nosocomial infection remained constant. 28 Apart from malnutrition, central venous catheter, and duration of mechanical ventilation, which are well-known variables associated with nosocomial infection, 27–30 we also identified unscheduled surgery as an important risk factor. This finding is in agreement with the results of the European Prevalence of Infection in Intensive Care study. 28 Finally, in this particular patient population, age more than 75 yr is not a determinant risk factor of nosocomial infection when other important variables are taken into account.
Many investigations, mainly performed in medical ICU, have assessed quantitatively the impact and influence of age on recovery from critical illness. Many of these studies show that mortality in the ICU increases with age. 1,4–6,10 Four large studies have identified age as an independent predictor of ICU death. 5,6,28,37 However, other factors, including primary disease, associated morbidities and complications, and severity of illness, significantly influence outcome, and age alone therefore cannot be used to determine mortality. 5,6 Katzman McClish et al.  4 found that ICU mortality increased from 18% in patients aged 55 to 64 yr to 25% in patients 75 yr of age or more, but this difference disappeared when logistic regression was used to control for demographic and diagnostic variables. However, some studies were not able to document an increased mortality rate in ICU among elderly, 11–13,26 suggesting that age is not an independent predictor of death in the ICU. Severity of illness, as assessed by the APACHE II score, was a better predictor of survival than age. Thus, in the APACHE scoring system, age accounts for a minority of the total explanatory power of the tool. 14 Other independent variables appear to be more important than age to predict ICU death in a surgical ICU: ASA score more than II, 38 sepsis, 28,37 need for mechanical ventilation more than 24 hr, 15 nosocomial infection, 7,27,28,33 and APACHE II score. 14 In addition, age more than 75 yr does not predict in-hospital mortality for patients who have been in a surgical ICU, suggesting the absence of bias concerning ICU discharge of elderly patients.
One concern is that selection bias may have influenced our results. Although we were unable to control bias among surgeons or attending physicians not to perform certain major surgeries on elderly patients, several points nevertheless make our population sample representative and valid for analysis. First, concerning the presence of comorbidities, the proportion of elderly patients classified with an ASA status of more than II (37%) was similar to that of a large French survey. 39 Second, the respective percentages of surgical procedures leading to ICU admission reflect the marked changes in the types and number of surgical procedures performed during the last decade, 39 as in other countries. 40 Third, elderly patients do not appear to have had fewer major surgeries than nonelderly patients (table can be found on the Anesthesiology Web site). Specifically, when focusing on the three main types of major surgery (orthopedic, vascular, and digestive), elderly patients had equal or higher percentages compared with younger patients.
An important issue concerns the extent of resources that are provided to elderly patients who gain admission to an ICU. Recently, Castillo-Lorente et al.  1 and Hamel et al.  , 41 in large prospective multicenter studies, reported a reduction in therapeutic activity concerning patients more than 75 yr of age. Accordingly, similar or slightly lower levels of therapeutic activity could be interpreted as a limitation of the use of therapeutic resources. However, it seems that the survival disadvantage experienced by seriously ill elderly patients is not explained by the less aggressive treatment they receive. 42 For example, mechanical ventilation is extensively used and may be an important variable in assessing the appropriateness of ICU resource use in elderly patients. Many recent studies have shown that the percentage of mechanically ventilated elderly patients was similar to that in younger patients, 6,12 and the duration of mechanical ventilation was also similar. 12,13,26 These results and the results of our study argue against the use of chronological age to decide whether to use mechanical ventilation.
The objective of ICU treatment is not simply to keep the patient alive in the hospital. Only 24% of older patients from our study were able to return home, which is at variance with the 40% found in the study by Chelluri et al.  12 in a medical ICU. This may reflect the need for physical rehabilitation after surgical procedures rather than a definite worsened status. Thus, in the study by Mahul et al.  25 that comprised 25% of admissions after elective surgery, 91% of elderly patients were able to return home 6 months after hospital discharge. After hospital discharge, most elderly patients returned to their prehospital functional status. 6,10,12,25 Elderly patients, requiring intensive care, described their quality of life as adequate and were willing to receive intensive care again, if necessary. 12 
In conclusion, in patients referred to a surgical ICU after a surgical procedure, age more than 75 yr by itself does not appear to be a significant predictor of ICU-acquired nosocomial infection or mortality rate during the ICU stay. Although age is a convenient and intuitively plausible marker for allocating medical resources, the data presented here do not support a policy to limit admissions to ICU or resource use solely on the basis of chronological age, even in patients older than 75 yr.
Table. Appendix:Omega Scoring System
Image not available
Table. Appendix:Omega Scoring System
×
Table 1. Characteristics of the Study Population
Image not available
Table 1. Characteristics of the Study Population
×
Table 2. Nosocomial Infections and Selected Risk Factors*
Image not available
Table 2. Nosocomial Infections and Selected Risk Factors*
×
Table 3. Outcome Variables and Therapeutic Activity*
Image not available
Table 3. Outcome Variables and Therapeutic Activity*
×
Table. Appendix:Omega Scoring System
Image not available
Table. Appendix:Omega Scoring System
×