Assessing the impact of real-time random safety audits through full propensity score matching on reliable data from the clinical information system.

BACKGROUND: Evidence-based care processes are not always applied at the bedside in critically ill patients. Numerous studies have assessed the impact of checklists and related strategies on the process of care and patient outcomes. We aimed to evaluate the effects of real-time random safety audits on process-of-care and outcome variables in critical care patients. METHODS: This prospective study used data from the clinical information system to evaluate the impact of real-time random safety audits targeting 32 safety measures in two intensive care units during a 9-month period. We compared endpoints between patients attended with safety audits and those not attended with safety audits. The primary endpoint was mortality, measured by Cox hazard regression after full propensity-score matching. Secondary endpoints were the impact on adherence to process-of-care measures and on quality indicators. RESULTS: We included 871 patients; 228 of these were attended in ≥ 1 real-time random safety audits. Safety audits were carried out on 390 patient-days; most improvements in the process of care were observed in safety measures related to mechanical ventilation, renal function and therapies, nutrition, and clinical information system. Although the group of patients attended in safety audits had more severe disease at ICU admission [APACHE II score 21 (16-27) vs. 20 (15-25), p = 0.023]; included a higher proportion of surgical patients [37.3 % vs. 26.4 %, p = 0.003] and a higher proportion of mechanically ventilated patients [72.8 % vs. 40.3 %, p < 0.001]; averaged more days on mechanical ventilation, central venous catheter, and urinary catheter; and had a longer ICU stay [12.5 (5.5-23.3) vs. 2.9 (1.7-5.9), p < 0.001], ICU mortality did not differ significantly between groups (19.3 % vs. 18.8 % in the group without safety rounds). After full propensity-score matching, Cox hazard regression analysis showed real-time random safety audits were associated with a lower risk of mortality throughout the ICU stay (HR 0.31; 95 %CI 0.20-0.47). CONCLUSIONS: Real-time random safety audits are associated with a reduction in the risk of ICU mortality. Exploiting data from the clinical information system is useful in assessing the impact of them on the care process, quality indicators, and mortality.

Automatic generation of minimum dataset and quality indicators from data collected routinely by the clinical information system in an intensive care unit.

BACKGROUND: Quality indicators (QIs) are being increasingly used in medicine to compare and improve the quality of care delivered. The feasibility of data collection is an important prerequisite for QIs. Information technology can improve efforts to measure processes and outcomes. In intensive care units (ICU), QIs can be automatically measured by exploiting data from clinical information systems (CIS). OBJECTIVE: To describe the development and application of a tool to automatically generate a minimum dataset (MDS) and a set of ICU quality metrics from CIS data. METHODS: We used the definitions for MDS and QIs proposed by the Spanish Society of Critical Care Medicine and Coronary Units. Our tool uses an extraction, transform, and load process implemented with Python to extract data stored in various tables in the CIS database and create a new associative database. This new database is uploaded to Qlik Sense, which constructs the MDS and calculates the QIs by applying the required metrics. The tool was tested using data from patients attended in a 30-bed polyvalent ICU during a six-year period. RESULTS: We describe the definitions and metrics, and we report the MDS and QI measurements obtained through the analysis of 4546 admissions. The results show that our ICU's performance on the QIs analyzed meets the standards proposed by our national scientific society. CONCLUSIONS: This is the first step toward using a tool to automatically obtain a set of actionable QIs to monitor and improve the quality of care in ICUs, eliminating the need for professionals to enter data manually, thus saving time and ensuring data quality.

Validation of the ICU-DaMa tool for automatically extracting variables for minimum dataset and quality indicators: The importance of data quality assessment.

BACKGROUND: Big data analytics promise insights into healthcare processes and management, improving outcomes while reducing costs. However, data quality is a major challenge for reliable results. Business process discovery techniques and an associated data model were used to develop data management tool, ICU-DaMa, for extracting variables essential for overseeing the quality of care in the intensive care unit (ICU). OBJECTIVE: To determine the feasibility of using ICU-DaMa to automatically extract variables for the minimum dataset and ICU quality indicators from the clinical information system (CIS). METHODS: The Wilcoxon signed-rank test and Fisher's exact test were used to compare the values extracted from the CIS with ICU-DaMa for 25 variables from all patients attended in a polyvalent ICU during a two-month period against the gold standard of values manually extracted by two trained physicians. Discrepancies with the gold standard were classified into plausibility, conformance, and completeness errors. RESULTS: Data from 149 patients were included. Although there were no significant differences between the automatic method and the manual method, we detected differences in values for five variables, including one plausibility error and two conformance and completeness errors. Plausibility: 1) Sex, ICU-DaMa incorrectly classified one male patient as female (error generated by the Hospital's Admissions Department). Conformance: 2) Reason for isolation, ICU-DaMa failed to detect a human error in which a professional misclassified a patient's isolation. 3) Brain death, ICU-DaMa failed to detect another human error in which a professional likely entered two mutually exclusive values related to the death of the patient (brain death and controlled donation after circulatory death). Completeness: 4) Destination at ICU discharge, ICU-DaMa incorrectly classified two patients due to a professional failing to fill out the patient discharge form when thepatients died. 5) Length of continuous renal replacement therapy, data were missing for one patient because the CRRT device was not connected to the CIS. CONCLUSIONS: Automatic generation of minimum dataset and ICU quality indicators using ICU-DaMa is feasible. The discrepancies were identified and can be corrected by improving CIS ergonomics, training healthcare professionals in the culture of the quality of information, and using tools for detecting and correcting data errors.

Impact of random safety analyses on structure, process and outcome indicators: multicentre study.

BACKGROUND: To assess the impact of a real-time random safety tool on structure, process and outcome indicators. METHODS: Prospective study conducted over a period of 12 months in two adult patient intensive care units. Safety rounds were conducted three days a week ascertaining 37 safety measures (grouped into 10 blocks). In each round, 50% of the patients and 50% of the measures were randomized. The impact of this safety tool was analysed on indicators of structure (safety culture, healthcare protocols), process (improvement proportion related to tool application, IPR) and outcome (mortality, average stay, rate of catheter-related bacteraemias and rate of ventilator-associated pneumonia, VAP). RESULTS: A total of 1214 patient-days were analysed. Structure indicators: the use of the safety tool was associated with an increase in the safety climate and the creation/modification of healthcare protocols (sedation/analgesia and weaning). Process indicators: Twelve of the 37 measures had an IPR > 10%; six showed a progressive decrease in the IPR over the study period. Nursing workloads and patient severity on the day of analysis were independently associated with a higher IPR in half of the blocks of variables. Outcome indicators: A significant decrease in the rate of VAP was observed. CONCLUSIONS: The real-time random safety tool improved the care process and adherence to clinical practice guidelines and was associated with an improvement in structure, process and outcome indicators.

Genome-wide association study of survival from sepsis due to pneumonia: an observational cohort study.

BACKGROUND: Sepsis continues to be a major cause of death, disability, and health-care expenditure worldwide. Despite evidence suggesting that host genetics can influence sepsis outcomes, no specific loci have yet been convincingly replicated. The aim of this study was to identify genetic variants that influence sepsis survival. METHODS: We did a genome-wide association study in three independent cohorts of white adult patients admitted to intensive care units with sepsis, severe sepsis, or septic shock (as defined by the International Consensus Criteria) due to pneumonia or intra-abdominal infection (cohorts 1-3, n=2534 patients). The primary outcome was 28 day survival. Results for the cohort of patients with sepsis due to pneumonia were combined in a meta-analysis of 1553 patients from all three cohorts, of whom 359 died within 28 days of admission to the intensive-care unit. The most significantly associated single nucleotide polymorphisms (SNPs) were genotyped in a further 538 white patients with sepsis due to pneumonia (cohort 4), of whom 106 died. FINDINGS: In the genome-wide meta-analysis of three independent pneumonia cohorts (cohorts 1-3), common variants in the FER gene were strongly associated with survival (p=9·7 × 10(-8)). Further genotyping of the top associated SNP (rs4957796) in the additional cohort (cohort 4) resulted in a combined p value of 5·6 × 10(-8) (odds ratio 0·56, 95% CI 0·45-0·69). In a time-to-event analysis, each allele reduced the mortality over 28 days by 44% (hazard ratio for death 0·56, 95% CI 0·45-0·69; likelihood ratio test p=3·4 × 10(-9), after adjustment for age and stratification by cohort). Mortality was 9·5% in patients carrying the CC genotype, 15·2% in those carrying the TC genotype, and 25·3% in those carrying the TT genotype. No significant genetic associations were identified when patients with sepsis due to pneumonia and intra-abdominal infection were combined. INTERPRETATION: We have identified common variants in the FER gene that associate with a reduced risk of death from sepsis due to pneumonia. The FER gene and associated molecular pathways are potential novel targets for therapy or prevention and candidates for the development of biomarkers for risk stratification. FUNDING: European Commission and the Wellcome Trust.

Toll-like receptors 2 and 4 contribute to sepsis-induced depletion of spleen dendritic cells.

Depletion of dendritic cells (DC) in secondary lymphoid organs is a hallmark of sepsis-induced immune dysfunction. In this setting, we investigated if Toll-like receptor (TLR)-dependent signaling might modulate the maturation process and the survival of DC. Using a model of sublethal polymicrobial sepsis induced by cecal ligation and puncture, we investigated the quantitative and functional features of spleen DC in wild-type, TLR2(-/-), TLR4(-/-), and TLR2(-/-) TLR4(-/-) mice. By 24 h, a decrease in the relative percentage of CD11c(high) spleen DC occurred in wild-type mice but was prevented in TLR2(-/-), TLR4(-/-), and TLR2(-/-) TLR4(-/-) mice. In wild-type mice, sepsis dramatically affected both CD11c(+) CD8alpha(+) and CD11c(+) CD8alpha(-) subsets. In all three types of knockout mice studied, the CD11c(+) CD8alpha(+) subset followed a depletion pattern similar to that for wild-type mice. In contrast, the loss of CD11c(+) CD8alpha(-) cells was attenuated in TLR2(-/-) and TLR4(-/-) mice and completely prevented in TLR2(-/-) TLR4(-/-) mice. Accordingly, apoptosis of spleen DC was increased in septic wild-type mice and inhibited in knockout mice. In addition we characterized the functional features of spleen DC obtained from septic mice. As shown by increased expression of major histocompatibility complex class II and CD86, polymicrobial sepsis induced maturation of DC, with subsequent increased capacity to prime T lymphocytes, similarly in wild-type and knockout mice. In response to CpG DNA stimulation, production of interleukin-12 was equally impaired in DC obtained from wild-type and knockout septic mice. In conclusion, although dispensable for the DC maturation process, TLR2 and TLR4 are involved in the mechanisms leading to depletion of spleen DC following polymicrobial sepsis.

PAI-1 gene: pharmacogenetic association of 4G/4G genotype with bleeding after cardiac surgery--pilot study.

BACKGROUND AND OBJECTIVE: To investigate whether the 4G/4G genotype of the PAI-1 gene is associated with bleeding after cardiac surgery and whether it may influence the use of antifibrinolytic drugs. METHODS: After a case-control association study to compare the distribution of genotypes of the 4G/5G polymorphism of the PAI-1 gene (4G/4G, 4G/5G and 5G/5G) between cardiac surgery patients (n = 260) and nonhospitalized age-matched controls (n = 111), we have evaluated the possible association of genotype homozygous 4G/4G (considered procoagulant) in two cohorts of cardiac surgery patients (treated with aprotinin or tranexamic acid) with postoperative bleeding and transfusion requirements. Chest tube output was measured at 6 h and 24 h and then total blood output. Genotypes were typed using restriction fragment length polymorphism analysis. RESULTS: The PAI-1 4G/4G genotype was not associated with bleeding except in the subgroup of patients treated with aprotinin in whom blood loss was significantly lower than in nonhomozygous 4G/4G patients at 6 h and 24 h [mean 135.9 ml (SD 101.8 ml) vs. mean 227.6 ml (SD 218.2 ml), P < 0.05; mean 314.5 ml (SD 180.3) vs. mean 482.7 ml (SD 349.8), P < 0.05, respectively]. Moreover, in homozygous 4G/4G patients, aprotinin was independently associated with lower total blood loss and also tended to require less transfusion (26.3 vs. 47.2%; 95% confidence interval, 0.3-2.7; P = 0.2). Only the European system of cardiac-operative risk evaluation score of at least 6 and therapy with platelet antiaggregants were associated with bleeding in the general patient population. CONCLUSION: The 4G/4G genotype of the PAI-1 gene was associated with less bleeding after cardiac surgery only in the subgroup of patients treated with aprotinin.

Risk factors for ventilator-associated pneumonia by Pseudomonas aeruginosa in presence of recent antibiotic exposure.

BACKGROUND: To facilitate the decision-making process for therapy and prevention of ventilator-associated pneumonia (VAP) in patients undergoing recent antibiotic exposure, this study investigated whether the development of VAP episodes caused by Pseudomonas aeruginosa or other pathogens are related to different risk factors, thereby distinguishing two risk population for this serious complication. METHODS: A 5-year retrospective case-control observational study was conducted. Cases of VAP caused by P. aeruginosa were compared with those caused by other pathogens. Univariate and multivariate analysis was performed using SPSS 11.0 software (SPSS Inc., Chicago, IL). RESULTS: Two groups were identified: P. aeruginosa (group P) was isolated in 58 (63.7%) episodes, and 33 episodes served as controls (group C), after a median of 12 days (interquartile range, 4-28 days) and 9 days (interquartile range, 3-12.5 days) of mechanical ventilation, respectively. P. aeruginosa was identified in 34.7% of episodes with early-onset pneumonia and in 73.5% with late-onset pneumonia. In a logistic regression analysis, P. aeruginosa was independently associated with duration of stay of 5 days or longer (relative risk = 3.59; 95% confidence interval, 1.04-12.35) and absence of coma (relative risk = 8.36; 95% confidence interval, 2.68-26.09). Risk for pathogens different from P. aeruginosa (group C) in early-onset pneumonia associated with coma was estimated to be 87.5%. CONCLUSIONS: Risk factors in episodes under recent antibiotic treatment caused by P. aeruginosa or other microorganism are not the same, a fact that could have implications for preventive and therapeutic approaches for this infection.

Association between 4G/5G polymorphism of the plasminogen activator inhibitor 1 gene with stroke or encephalopathy after cardiac surgery.

OBJECTIVE: To evaluate whether the 4G/5G polymorphism of the plasminogen activator inhibitor 1 gene increases the risk of thromboembolic neurological complications, and, if so, whether these complications lengthen the period of mechanical ventilation and hospital stay. DESIGN: Prospective, case-control study in a 14 bed surgical intensive care unit of a university hospital. PATIENTS: 260 consecutive patients who underwent cardiac surgery with cardiopulmonary bypass and 111 controls. INTERVENTIONS: DNA was isolated and 4G/5G polymorphism was typed using RFLP methodology. MEASUREMENTS AND RESULTS: Genetic analysis revealed 4G/5G in 131 patients (50.4%), 5G/5G genotype in 82 (31.5%), and 4G/4G in only 47 (18.1%). Prevalence of neurological complications was 20.8% (n=54) (stroke 5.4%, n=14; encephalopathy 15.4%, n=40]. A trend towards higher risk of developing stroke (8.5% vs. 4.7%, RR 1.9) and a significant twofold increase in encephalopathy (27.7% vs. 12.7%; RR 2.6) was documented in 4G/4G carriers. Multivariate analysis showed that development of stroke or encephalopathy was independently associated with prolonged mechanical ventilation (OR 20), and that neurological complication (OR 2.4) and 4G/4G genotype (OR 2.6) were independently associated with hospital stay of 2 weeks or longer. CONCLUSIONS: The 4G/4G genotype can increase the risk of thromboembolic neurological complications after cardiac surgery with cardiopulmonary by-pass. The neurological complications result in longer time on ventilator and longer hospital stay.

Pneumonia caused by oxacillin-resistant Staphylococcus aureus treated with glycopeptides.

OBJECTIVE: To determine whether ventilator-associated pneumonia caused by oxacillin-resistant Staphylococcus aureus (VAP-ORSA) treated with glycopeptides is associated with an increased mortality rate. DESIGN: Retrospective matched cohort study. SETTING: Four intensive care units in teaching hospitals. PATIENTS: Seventy-five patients were matched to 75 controls. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: All adult intensive care unit patients with microbiologically documented VAP-ORSA were matched to intubated controls who did not develop VAP-ORSA, based on disease severity (Acute Physiology and Chronic Health Evaluation II score) at admission (+/-3 points), diagnostic category, and length of stay before pneumonia onset. Population characteristics and intensive care unit mortality rates of patients with VAP-ORSA and their controls without pneumonia were compared. Attributable mortality was determined by subtracting the crude mortality rate of controls from the crude mortality rate of VAP-ORSA patients. Thirty-six of the 75 matched VAP-ORSA patients died, representing a crude mortality rate of 48%, whereas 19 of the 75 controls died, a crude mortality rate of 25.3% (p < .01). Excess mortality was estimated to be 22.7% (95% confidence interval, 2.4-42.9%). Median length of intensive care unit stay in the surviving pairs was 33 days (interquartile range, 25-75%: 25-45 days) for VAP-ORSA patients and 21 days (interquartile range, 25-75%: 15-34.75 days) days for controls (p = .054). CONCLUSIONS: Despite appropriate glycopeptide therapy, there is an increased attributable mortality for pneumonia by ORSA, after careful adjustment for disease severity and diagnostic category.

Clinical approach to the patient with suspected ventilator-associated pneumonia.

Management of ventilator-associated pneumonia needs to balance the avoidance of unnecessary antibiotic overuse with the provision of adequate initial empiric therapy. A clinical diagnosis based on new pulmonary opacity and purulent respiratory secretions plus other signs of inflammation is valuable in screening for patients with suspected ventilator-associated pneumonia. A rational strategy starts with immediate initiation of adequate antibiotics and collection of respiratory secretions to evaluate the causative organism. As a minimum, an endotracheal aspirate with direct staining and quantitative cultures should be obtained. Overall, the need to choose adequate antibiotics correctly and expeditiously calls for the use of broad-spectrum antibiotics, but the choice should be narrowed quickly in the light of microbiologic information. However, some patients (those who develop an infection within 5 days of hospitalization, those without recent antibiotic exposure, and those without hospitalization in the past 3 months) are at low risk of infection by resistant organisms. In that subset, adequate initial selection could be a non-pseudomonal third-generation cephalosporin, since antibiotics should target usual community-acquired organisms in addition to some Enterobacteriaceae and Staphylococcus aureus. Coverage of methicillin-resistant S. aureus should be limited only to intensive care units with concomitant index cases and to patients under antibiotic exposure. Patients at risk of Pseudomonas aeruginosa (e.g., 1 week of prior hospitalization or chronic obstructive pulmonary disease) require initial use of a combination of piperacilin/tazobactam and ciprofloxacin, or amikacin plus imipenem, meropenem, or an antipseudomonal cephalosporin. If risk of Acinetobacter baumannii exists, one of these agents should be a carbapenem. After 48 hours of therapy, each patient should be re-evaluated based mainly on resolution of hypoxemia and fever plus the initial microbiologic information. Whereas broad-spectrum therapy is initially warranted in many patients, this treatment may be narrowed considerably as culture results identify the causative organism and its sensitivity. Recent data suggest that reducing overall treatment duration to a maximum of 1 week is safe, effective and is less likely to promote the growth of resistant organisms in patients who are clinically improving. Optimal management should be based on a strategy combining early high doses of an effective agent for a short period of time, which is then simplified in the light of microbiologic information.

Clinical resolution in patients with suspicion of ventilator-associated pneumonia: a cohort study comparing patients with and without acute respiratory distress syndrome.

OBJECTIVES: To determine the pattern of resolution of classic infectious and respiratory variables in patients with ventilator-associated pneumonia (VAP) and appropriate empirical therapy, depending on the presence of acute respiratory distress syndrome (ARDS). A secondary objective was to identify clinical variables that might be useful for monitoring response to therapy. DESIGN: Prospective, observational cohort study. SETTING: Medical-surgical intensive care unit. PATIENTS: Seventy-five episodes of VAP without ARDS were identified and compared with 20 episodes with ARDS at VAP onset. Six episodes were excluded due to in vitro resistance to the initial antibiotic choice and six due to death in the first 72 hrs. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Resolution of fever, Pao2/Fio2 >250 mm Hg, and white blood cell count in episodes of VAP were present in 73.3%, 74.7%, and 53.3% of patients after 3 days of therapy. Indeed, >50% of episodes with the absence of ARDS presented resolution of fever and Pao2/Fio2 >250 within the first day of therapy. In contrast, resolution of radiologic opacities and clearance of secretions (median of 14 and 6 days of resolution) were late events. In patients with ARDS, resolution of fever remained the earliest variable. However, similar to Pao2/Fio2 250 and white blood cell count, fever showed a significantly worse pattern after 3 days of therapy: 45%, 15% and 25%, respectively. Radiologic resolution was an extremely poor indicator, being present in only 10% of ARDS patients after 15 days of follow-up. Failure to improve after 48 hrs of therapy was documented in 65% of ARDS patients and 14.7% of controls (p < .05). CONCLUSIONS: Measures of oxygenation and core temperature can help physicians to individualize and shorten the duration of antibiotic therapy in VAP episodes. ARDS patients with VAP take twice as long to resolve fever, whereas hypoxemia should be ignored in defining resolution in this subset.

Pneumonia in head-injured and severe trauma patients.

Trauma is the leading cause of death in young people. Trauma deaths have a classic trimodal distribution; in late death (3 days to 3 weeks postinjury), infection is the principal cause. Because pneumonia is a major cause of morbidity in trauma patients, early identification of subjects at a high risk of developing nosocomial pneumonia (NP) can reduce morbidity and costs. Methicillin-sensitive Staphylococcus aureus (MSSA) is the predominant pathogen in multiple-trauma patients in coma, and nasal MSSA colonization at time of severe injury may increase the risk of MSSA pneumonia. In the remaining patients, gram-negative bacilli are responsible for the majority of cases. Prolonged mechanical ventilation, continuous enteral feeding, and craniotomy are risk factors for NP in trauma patients. Diagnosis of NP in these patients is difficult because radiographic infiltrates may not highlight any infection. In coma patients, coverage with a beta-lactam active against MSSA is mandatory. Variations in organisms and sensitivities across intensive care units due to differences in demographic characteristics or comorbidities should be considered.