Peter Dodek

Intensive versus conventional glucose control in critically ill patients

BACKGROUND The optimal target range for blood glucose in critically ill patients remains unclear. METHODS Within 24 hours after admission to an intensive care unit (ICU), adults who were expected to require treatment in the ICU on 3 or more consecutive days were randomly assigned to undergo either intensive glucose control, with a target blood glucose range of 81 to 108 mg per deciliter (4.5 to 6.0 mmol per liter), or conventional glucose control, with a target of 180 mg or less per deciliter (10.0 mmol or less per liter). We defined the primary end point as death from any cause within 90 days after randomization. RESULTS Of the 6104 patients who underwent randomization, 3054 were assigned to undergo intensive control and 3050 to undergo conventional control; data with regard to the primary outcome at day 90 were available for 3010 and 3012 patients, respectively. The two groups had similar characteristics at baseline. A total of 829 patients (27.5%) in the intensive-control group and 751 (24.9%) in the conventional-control group died (odds ratio for intensive control, 1.14; 95% confidence interval, 1.02 to 1.28; P=0.02). The treatment effect did not differ significantly between operative (surgical) patients and nonoperative (medical) patients (odds ratio for death in the intensive-control group, 1.31 and 1.07, respectively; P=0.10). Severe hypoglycemia (blood glucose level, < or = 40 mg per deciliter [2.2 mmol per liter]) was reported in 206 of 3016 patients (6.8%) in the intensive-control group and 15 of 3014 (0.5%) in the conventional-control group (P<0.001). There was no significant difference between the two treatment groups in the median number of days in the ICU (P=0.84) or hospital (P=0.86) or the median number of days of mechanical ventilation (P=0.56) or renal-replacement therapy (P=0.39). CONCLUSIONS In this large, international, randomized trial, we found that intensive glucose control increased mortality among adults in the ICU: a blood glucose target of 180 mg or less per deciliter resulted in lower mortality than did a target of 81 to 108 mg per deciliter. (ClinicalTrials.gov number, NCT00220987.)

Canadian clinical practice guidelines for nutrition support in mechanically ventilated, critically ill adult patients

OBJECTIVE This study was conducted to develop evidence-based clinical practice guidelines for nutrition support (ie, enteral and parenteral nutrition) in mechanically ventilated critically ill adults. OPTIONS The following interventions were systematically reviewed for inclusion in the guidelines: enteral nutrition (EN) versus parenteral nutrition (PN), early versus late EN, dose of EN, composition of EN (protein, carbohydrates, lipids, immune-enhancing additives), strategies to optimize delivery of EN and minimize risks (ie, rate of advancement, checking residuals, use of bedside algorithms, motility agents, small bowel versus gastric feedings, elevation of the head of the bed, closed delivery systems, probiotics, bolus administration), enteral nutrition in combination with supplemental PN, use of PN versus standard care in patients with an intact gastrointestinal tract, dose of PN and composition of PN (protein, carbohydrates, IV lipids, additives, vitamins, trace elements, immune enhancing substances), and the use of intensive insulin therapy. OUTCOMES The outcomes considered were mortality (intensive care unit [ICU], hospital, and long-term), length of stay (ICU and hospital), quality of life, and specific complications. EVIDENCE We systematically searched MEDLINE and CINAHL (cumulative index to nursing and allied health), EMBASE, and the Cochrane Library for randomized controlled trials and meta-analyses of randomized controlled trials that evaluated any form of nutrition support in critically ill adults. We also searched reference lists and personal files, considering all articles published or unpublished available by August 2002. Each included study was critically appraised in duplicate using a standard scoring system. VALUES For each intervention, we considered the validity of the randomized trials or meta-analyses, the effect size and its associated confidence intervals, the homogeneity of trial results, safety, feasibility, and the economic consequences. The context for discussion was mechanically ventilated patients in Canadian ICUs. BENEFITS, HARMS, AND COSTS The major potential benefit from implementing these guidelines is improved clinical outcomes of critically ill patients (reduced mortality and ICU stay). Potential harms of implementing these guidelines include increased complications and costs related to the suggested interventions. SUMMARIES OF EVIDENCE AND RECOMMENDATIONS: When considering nutrition support in critically ill patients, we strongly recommend that EN be used in preference to PN. We recommend the use of a standard, polymeric enteral formula that is initiated within 24 to 48 hours after admission to ICU, that patients be cared for in the semirecumbent position, and that arginine-containing enteral products not be used. Strategies to optimize delivery of EN (starting at the target rate, use of a feeding protocol using a higher threshold of gastric residuals volumes, use of motility agents, and use of small bowel feeding) and minimize the risks of EN (elevation of the head of the bed) should be considered. Use of products with fish oils, borage oils, and antioxidants should be considered for patients with acute respiratory distress syndrome. A glutamine-enriched formula should be considered for patients with severe burns and trauma. When initiating EN, we strongly recommend that PN not be used in combination with EN. When PN is used, we recommend that it be supplemented with glutamine, where available. Strategies that maximize the benefit and minimize the risks of PN (hypocaloric dose, withholding lipids, and the use of intensive insulin therapy to achieve tight glycemic control) should be considered. There are insufficient data to generate recommendations in the following areas: use of indirect calorimetry; optimal pH of EN; supplementation with trace elements, antioxidants, or fiber; optimal mix of fats and carbohydrates; use of closed feeding systems; continuous versus bolus feedings; use of probiotics; type of lipids; and mode of lipid delivery. VALIDATION This guideline was peer-reviewed and endorsed by official representatives of the Canadian Critical Care Society, Canadian Critical Care Trials Group, Dietitians of Canada, Canadian Association of Critical Care Nurses, and the Canadian Society for Clinical Nutrition. SPONSORS This guideline is a joint venture of the Canadian Critical Care Society, the Canadian Critical Trials Group, the Canadian Society for Clinical Nutrition, and Dietitians of Canada. The Canadian Critical Care Society and the Institute of Nutrition, Metabolism, and Diabetes of the Canadian Institutes of Health Research provided funding for development of this guideline.

Critically Ill Patients With 2009 Influenza A(H1N1) Infection in Canada

CONTEXT Between March and July 2009, the largest number of confirmed cases of 2009 influenza A(H1N1) infection occurred in North America. OBJECTIVE To describe characteristics, treatment, and outcomes of critically ill patients in Canada with 2009 influenza A(H1N1) infection. DESIGN, SETTING, AND PATIENTS A prospective observational study of 168 critically ill patients with 2009 influenza A(H1N1) infection in 38 adult and pediatric intensive care units (ICUs) in Canada between April 16 and August 12, 2009. MAIN OUTCOME MEASURES The primary outcome measures were 28-day and 90-day mortality. Secondary outcomes included frequency and duration of mechanical ventilation and duration of ICU stay. RESULTS Critical illness occurred in 215 patients with confirmed (n = 162), probable (n = 6), or suspected (n = 47) community-acquired 2009 influenza A(H1N1) infection. Among the 168 patients with confirmed or probable 2009 influenza A(H1N1), the mean (SD) age was 32.3 (21.4) years; 113 were female (67.3%) and 50 were children (29.8%). Overall mortality among critically ill patients at 28 days was 14.3% (95% confidence interval, 9.5%-20.7%). There were 43 patients who were aboriginal Canadians (25.6%). The median time from symptom onset to hospital admission was 4 days (interquartile range [IQR], 2-7 days) and from hospitalization to ICU admission was 1 day (IQR, 0-2 days). Shock and nonpulmonary acute organ dysfunction was common (Sequential Organ Failure Assessment mean [SD] score of 6.8 [3.6] on day 1). Neuraminidase inhibitors were administered to 152 patients (90.5%). All patients were severely hypoxemic (mean [SD] ratio of Pao(2) to fraction of inspired oxygen [Fio(2)] of 147 [128] mm Hg) at ICU admission. Mechanical ventilation was received by 136 patients (81.0%). The median duration of ventilation was 12 days (IQR, 6-20 days) and ICU stay was 12 days (IQR, 5-20 days). Lung rescue therapies included neuromuscular blockade (28% of patients), inhaled nitric oxide (13.7%), high-frequency oscillatory ventilation (11.9%), extracorporeal membrane oxygenation (4.2%), and prone positioning ventilation (3.0%). Overall mortality among critically ill patients at 90 days was 17.3% (95% confidence interval, 12.0%-24.0%; n = 29). CONCLUSION Critical illness due to 2009 influenza A(H1N1) in Canada occurred rapidly after hospital admission, often in young adults, and was associated with severe hypoxemia, multisystem organ failure, a requirement for prolonged mechanical ventilation, and the frequent use of rescue therapies.

Initiation of inappropriate antimicrobial therapy results in a fivefold reduction of survival in human septic shock.

OBJECTIVE Our goal was to determine the impact of the initiation of inappropriate antimicrobial therapy on survival to hospital discharge of patients with septic shock. METHODS The appropriateness of initial antimicrobial therapy, the clinical infection site, and relevant pathogens were retrospectively determined for 5,715 patients with septic shock in three countries. RESULTS Therapy with appropriate antimicrobial agents was initiated in 80.1% of cases. Overall, the survival rate was 43.7%. There were marked differences in the distribution of comorbidities, clinical infections, and pathogens in patients who received appropriate and inappropriate initial antimicrobial therapy (p < 0.0001 for each). The survival rates after appropriate and inappropriate initial therapy were 52.0% and 10.3%, respectively (odds ratio [OR], 9.45; 95% CI, 7.74 to 11.54; p < 0.0001). Similar differences in survival were seen in all major epidemiologic, clinical, and organism subgroups. The decrease in survival with inappropriate initial therapy ranged from 2.3-fold for pneumococcal infection to 17.6-fold with primary bacteremia. After adjustment for acute physiology and chronic health evaluation II score, comorbidities, hospital site, and other potential risk factors, the inappropriateness of initial antimicrobial therapy remained most highly associated with risk of death (OR, 8.99; 95% CI, 6.60 to 12.23). CONCLUSIONS Inappropriate initial antimicrobial therapy for septic shock occurs in about 20% of patients and is associated with a fivefold reduction in survival. Efforts to increase the frequency of the appropriateness of initial antimicrobial therapy must be central to efforts to reduce the mortality of patients with septic shock.

Evidence-based clinical practice guideline for the prevention of ventilator-associated pneumonia.

Critically ill patients in the intensive care unit (ICU) are at high risk for infections associated with increased morbidity, mortality, and health care costs (1-3). The overall infection rate in critically ill patients approaches 40% and may be as high as 50% or 60% in patients who remain in the ICU for more than 5 days (4, 5). Respiratory tract infections account for 30% to 60% of all such infections. The incidence of pneumonia acquired in the ICU ranges from 10% to 65% (6-11). Among patients at high risk for ventilator-associated pneumonia (VAP) are those who have chronic obstructive pulmonary disease, burns, neurosurgical conditions, the acute respiratory distress syndrome, and witnessed aspiration; those who are reintubated; and those who receive paralytic agents or enteral nutrition (12, 13). The attributable morbidity and mortality of VAP are clinically important. In a prospective, matched cohort study, patients with VAP remained in the ICU 4.3 days (95% CI, 1.5 to 7.0 days) longer than patients who did not have VAP and had a trend toward an increased risk for death (absolute risk increase, 5.8% [CI, 2.4% to 14.0%]) (14). Six other studies using a matching strategy found a prolonged length of ICU stay associated with VAP (range, 5 to 13 days) and attributable mortality ranging from an absolute risk increase of 0% to 50% (15-20). Therefore, strategies to decrease the incidence of VAP could decrease morbidity, mortality, and health care costs and improve patient safety. A survey of the use of VAP prevention strategies identified differences across countries (21). For example, changing the ventilator circuit for each new patient was reported more frequently by French ICU directors than those in Canada (21). This survey also showed that some effective strategies were used infrequently, suggesting inadequate translation of randomized trial results into practice. One potential catalyst for knowledge translation is an evidence-based clinical practice guideline. Therefore, a Joint Planning Group of the Canadian Critical Care Society and Canadian Critical Care Trials Group commissioned the development of an evidence-based clinical practice guideline for the prevention of VAP. In this paper, we describe the methods used to create the guideline and the recommendations generated. Methods The Joint Planning Group selected an 11-member VAP Prevention Guideline Panel made up of 9 intensivists from university-affiliated and community hospitals, an ICU nurse, and an ICU respiratory therapist. Panel members were experts in critical care medicine (n= 9), VAP (n= 4), evidence-based medicine (n= 4), and guideline development (n= 3). The context was mechanically ventilated adult patients cared for in the ICU. The target audience was ICU clinicians in university-affiliated and community hospitals. To identify potentially relevant evidence, we searched 3 bibliographic databases (MEDLINE, EMBASE, and the Cochrane Database of Systematic Reviews) to 1 April 2003 for randomized trials that evaluated interventions influencing VAP (Appendix). We had no language restrictions. We also reviewed personal files and practice guidelines on this subject previously published by the Centers for Disease Control and Prevention (22) and the American Thoracic Society (23). We included randomized trials and systematic reviews of randomized trials that 1) studied adult critically ill patients; 2) had VAP as an outcome; and 3) evaluated any of the following interventions: physical strategies (route of endotracheal intubation, systematic search for maxillary sinusitis, frequency of ventilator circuit changes, type of airway humidification, frequency of humidifier changes, endotracheal suctioning system, subglottic secretion drainage, chest physiotherapy, and tracheostomy timing), positional strategies (kinetic beds, semi-recumbent positioning, and prone positioning), and pharmacologic strategies (stress ulcer prophylaxis and prophylactic antibiotics, including selective decontamination of the digestive tract). Since study authors used various definitions of VAP, we used the definitions they provided. The most common definition was a new or persistent radiographic infiltrate plus fever, leukocytosis, change in the volume or color of sputum, or isolation of a pathogen. If available, histologic evidence of pneumonia was also used. A priori, we decided to review only systematic reviews of randomized clinical trials for antibiotic prophylaxis and only randomized clinical trials for all other topics. We excluded crossover and beforeafter studies. We also excluded randomized trials of ventilator weaning, including noninvasive mechanical ventilation, and nutritional interventions evaluating VAP because guidelines addressing these topics have recently been published (24, 25). In duplicate and independently, 3 pairs of panel members critically appraised each trial (26, 27) and systematic review (28). Each member of a pair compared his or her independent appraisal of a given trial or systematic review with that of the other member of the pair. For each randomized trial, we abstracted sample, allocation, intervention, co-interventions, exclusions after randomization, blinding of outcome assessment, definition of VAP, crude VAP events, relative risk for VAP, and other outcomes. For each intervention, we summarized the risk differences and calculated a pooled risk difference. For each systematic review, we abstracted number of trials, population, intervention, selection criteria, search strategy, validity assessment, method of pooling results, homogeneity assessment, VAP definition, pooled event rates, and other outcomes. Before the panel meeting, each pair of appraisers achieved consensus on the validity and results of the trials they reviewed. One month before the panel meeting, panel members received the evidence tables for review prepared by the 3 pairs of appraisers. A priori, panel members agreed to read all circulated documents and evidence tables in advance, to use levels of evidence to generate a status statement for each item, and to abide by the group process and consensus methods. The Canadian Critical Care Society appointed a chair to ensure that the panel achieved its objectives through group process (29). At the panel meeting, each member recorded any potential conflicts of interest (30). The pair of panel members responsible for critical appraisal of each intervention provided a structured written and oral presentation of the evidence. After the panel discussion, the initial evidence summary was revised if necessary. The panel members assigned levels of evidence, semi-quantitative scores to summarize the evidence and describe the intervention, and a status statement. We classified trials as level 1 if they had all of the following: concealed randomization, blinded outcome adjudication, an intention-to-treat analysis, and an explicit definition of VAP. Trials were classified as level 2 if any one of these characteristics was unfulfilled and as level 3 if allocation was not strictly randomized. We used a semi-quantitative score (0, 1, 2, or 3) to evaluate each intervention with respect to the validity of the randomized trials; the effect size of each intervention; the confidence intervals around the estimate of effect; the homogeneity of the trial results; and the safety, feasibility, and economic consequences of the intervention. The language of the status statement for each item was keyed to the levels of evidence and the semi-quantitative scores. We used the term recommended if there were no reservations about endorsing an intervention and the term considered if the evidence supported an intervention but there were minor uncertainties about the benefits, harms, or costs. No recommendation was made if evidence regarding an intervention was inadequate or if there were major uncertainties about the benefits, harms, or costs. After the panel meeting, the chair compiled the summaries and status statements and sent them to all panel members to check accuracy and clarity. In addition, the pairs of evidence appraisers wrote background documents for the interventions they appraised, including the rationale for each intervention, appraisal of randomized trials and systematic reviews, and harms and costs of the interventions. The chair and the writing committee organized the background documents, the evidence summaries, a table of the semi-quantitative scores, and the status statement for each item. We formatted the document with a structured abstract (31), a summary of the evidentiary basis for each recommendation, and a status statement for each item. We also created a quick reference guide. The draft guideline document was submitted for structured external review by the executives of the Canadian Critical Care Society and the Canadian Critical Care Trials Group and the respective executives of the Canadian Association of Critical Care Nurses, Canadian Society of Respiratory Therapists, Canadian Infectious Disease Society, and Canadian Thoracic Society. External reviewers were asked to critique whether the guideline was logical, clear, and practical and to critique the guideline development process. The panel revised the document on the basis of this feedback. The final guideline was returned to the external reviewers for further comments and official endorsement by their respective organizations. The final guideline was then piloted in 2 institutions. To record the agreement of each panel member with the final status statement for each item, we sent the final document to all panel members. Independently, blinded to each others ratings, panel members used a Likert scale from 1 to 9 that was anchored by disagree completely at the low end and agree completely at the high end. The panel will formally review and update this guideline every 2 years (32). The funding source played no role in study selection for this guideline and had no role in its development

Family satisfaction with care in the intensive care unit: results of a multiple center study.

Objective To determine the level of satisfaction of family members with the care that they and their critically ill relative received. Design Prospective cohort study. Setting Six university-affiliated intensive care units across Canada. Methods We administered a validated questionnaire to family members who made at least one visit to intensive care unit patients who received mechanical ventilation for >48 hrs. We obtained self-rated levels of satisfaction with 25 key aspects of care related to the overall intensive care unit experience, communication, and decision making. For family members of survivors, the questionnaire was administered while the patient was still in the hospital. For family members of nonsurvivors, the questionnaire was mailed out to the family member 3–4 wks after the patient’s death. Main Results A total of 891 family members received questionnaires; 624 were returned (70% response rate). The majority of respondents were satisfied with overall care and with overall decision making (mean ± sd item score, 84.3 ± 15.7 and 75.9 ± 26.4, respectively). Families reported the greatest satisfaction with nursing skill and competence (92.4 ± 14.0), the compassion and respect given to the patient (91.8 ± 15.4), and pain management (89.1 ± 16.7). They were least satisfied with the waiting room atmosphere (65.0 ± 30.6) and frequency of physician communication (70.7 ± 29.0). The variables significantly associated with overall satisfaction in a regression analysis were completeness of information received, respect and compassion shown to the patient and family member, and the amount of health care received. Satisfaction varied significantly across sites. Conclusions Most family members were highly satisfied with the care provided to them and their critically ill relative in the intensive care unit. Efforts to improve the nature of interactions and communication with families are likely to lead to improvements in satisfaction.

Early combination antibiotic therapy yields improved survival compared with monotherapy in septic shock: a propensity-matched analysis.

Background:Septic shock represents the major cause of infection-associated mortality in the intensive care unit. The possibility that combination antibiotic therapy of bacterial septic shock improves outcome is controversial. Current guidelines do not recommend combination therapy except for the express purpose of broadening coverage when resistant pathogens are a concern. Objective:To evaluate the therapeutic benefit of early combination therapy comprising at least two antibiotics of different mechanisms with in vitro activity for the isolated pathogen in patients with bacterial septic shock. Design:Retrospective, propensity matched, multicenter, cohort study. Setting:Intensive care units of 28 academic and community hospitals in three countries between 1996 and 2007. Subjects:A total of 4662 eligible cases of culture-positive, bacterial septic shock treated with combination or monotherapy from which 1223 propensity-matched pairs were generated. Measurements and Main Results:The primary outcome of study was 28-day mortality. Using a Cox proportional hazards model, combination therapy was associated with decreased 28-day mortality (444 of 1223 [36.3%] vs. 355 of 1223 [29.0%]; hazard ratio, 0.77; 95% confidence interval, 0.67-0.88; p = .0002). The beneficial impact of combination therapy applied to both Gram-positive and Gram-negative infections but was restricted to patients treated with &bgr;-lactams in combination with aminoglycosides, fluoroquinolones, or macrolides/clindamycin. Combination therapy was also associated with significant reductions in intensive care unit (437 of 1223 [35.7%] vs. 352 of 1223 [28.8%]; odds ratio, 0.75; 95% confidence interval, 0.63-0.92; p = .0006) and hospital mortality (584 of 1223 [47.8%] vs. 457 of 1223 [37.4%]; odds ratio, 0.69; 95% confidence interval, 0.59-0.81; p < .0001). The use of combination therapy was associated with increased ventilator (median and [interquartile range], 10 [0-25] vs. 17 [0-26]; p = .008) and pressor/inotrope-free days (median and [interquartile range], 23 [0-28] vs. 25 [0-28]; p = .007) up to 30 days. Conclusion:Early combination antibiotic therapy is associated with decreased mortality in septic shock. Prospective randomized trials are needed.

Comprehensive evidence-based clinical practice guidelines for ventilator-associated pneumonia : Prevention

BACKGROUND Ventilator-associated pneumonia (VAP) is an important cause of morbidity and mortality in ventilated critically ill patients. PURPOSE To develop evidence-based guidelines for the prevention of VAP. DATA SOURCES MEDLINE, EMBASE, CINAHL, and the Cochrane Database of Systematic Reviews and Register of Controlled Trials. STUDY SELECTION The authors systematically searched for all relevant randomized, controlled trials and systematic reviews on the topic of prevention of VAP in adults that were published from 1980 to October 1, 2006. DATA EXTRACTION Independently and in duplicate, the panel scored the internal validity of each trial. Effect size, confidence intervals, and homogeneity of the results were scored using predefined definitions. Scores for the safety, feasibility, and economic issues were assigned based on consensus of the guideline panel. LEVELS OF EVIDENCE The following statements were used: recommend, consider, do not recommend, and no recommendation due to insufficient or conflicting evidence. DATA SYNTHESIS To prevent VAP: We recommend: that the orotracheal route of intubation should be used for intubation; a new ventilator circuit for each patient; circuit changes if the circuit becomes soiled or damaged, but no scheduled changes; change of heat and moisture exchangers every 5 to 7 days or as clinically indicated; the use of a closed endotracheal suctioning system changed for each patient and as clinically indicated; subglottic secretion drainage in patients expected to be mechanically ventilated for more than 72 hours; head of bed elevation to 45 degrees (when impossible, as near to 45 degrees as possible should be considered). Consider: the use of rotating beds; oral antiseptic rinses. We do not recommend: use of bacterial filters; the use of iseganan We make no recommendations regarding: the use of a systematic search for sinusitis; type of airway humidification; timing of tracheostomy; prone positioning; aerosolized antibiotics; intranasal mupirocin; topical and/or intravenous antibiotics. CONCLUSION There are a growing number of evidence-based strategies for VAP prevention, which, if applied in practice, may reduce the incidence of this serious nosocomial infection.

Clinical practice guidelines for the use of noninvasive positive-pressure ventilation and noninvasive continuous positive airway pressure in the acute care setting

Over the past two decades, the use of noninvasive positive-pressure ventilation and noninvasive continuous positive airway pressure by mask has increased substantially for acutely ill patients. Initial case series and uncontrolled cohort studies that suggested benefit in selected patients[1][1]–[

Randomized trial of combination versus monotherapy for the empiric treatment of suspected ventilator-associated pneumonia.

Objective:To compare a strategy of combination therapy with a strategy of monotherapy with broad-spectrum antibiotics for suspected late ventilator-associated pneumonia. Design:Randomized trial. Setting:Twenty-eight intensive care units in Canada and the United States. Patients:The study included 740 mechanically ventilated patients who developed suspected ventilator-associated pneumonia after 96 hrs in the intensive care unit. Patients known to be colonized or infected with Pseudomonas or methicillin-resistant Staphylococcus aureus or who were immunocompromised were excluded from the study. Interventions:As initial unblinded therapy, patients were allocated to receive meropenem (1 g every 8 hrs) and ciprofloxacin (400 mg every 12 hrs) or meropenem alone. Before starting antibiotics, patients were also randomized to bronchoalveolar lavage with quantitative cultures or endotracheal aspirates. When culture results were available, physicians were encouraged to adjust antibiotics. Adequacy of antibiotics was defined as the organism present in the enrollment culture having in vitro susceptibility to one or more of the study antibiotics. Measurements and Main Results:Baseline characteristics and etiologies of ventilator-associated pneumonia were similar in the two groups. There was no difference in 28-day mortality between the combination and monotherapy groups (relative risk = 1.05, 95% confidence interval 0.78–1.42, p = .74). Duration of intensive care unit and hospital stay, clinical and microbiological treatment response, emergence of antibiotic-resistant bacteria, isolation of Clostridium difficile in stool, and fungal colonization were also similar in the two groups. In a subgroup of patients who had infection due to Pseudomonas species, Acinetobacter species, and multidrug-resistant Gram-negative bacilli at enrollment (n = 56), the adequacy of initial antibiotics (84.2% vs. 18.8%, p < .001) and microbiological eradication of infecting organisms (64.1% vs. 29.4%, p = .05) was higher in the combination group compared with the monotherapy group, but there were no differences in clinical outcomes. Conclusions:For critically ill patients who have suspected late ventilator-associated pneumonia and who are at low risk for difficult-to-treat Gram-negative bacteria, monotherapy is associated with similar outcomes compared with combination therapy. For those patients at high risk of difficult-to-treat Gram-negative bacteria, combination therapy is safe and may be associated with better microbiological and clinical outcomes.