Abstract
Rationale: Although intensive care units (ICUs) were created for patients with life-threatening illnesses, the ICU environment generates a high risk of iatrogenic events. Identifying medical errors (MEs) that serve as indicators for iatrogenic risk is crucial for purposes of reporting and prevention.
Objectives: We describe the selection of indicator MEs, the incidence of such MEs, and their relationship with mortality.
Methods: We selected indicator MEs using Delphi techniques. An observational prospective multicenter cohort study of these MEs was conducted from March 27 to April 3, 2006, in 70 ICUs; 16 (23%) centers were audited. Harm from MEs was collected using specific scales.
Measurements and Main Results: Fourteen types of MEs were selected as indicators; 1,192 MEs were reported for 1,369 patients, and 367 (26.8%) patients experienced at least 1 ME (2.1/1,000 patient-days). The most common MEs were insulin administration errors (185.9/1,000 d of insulin treatment). Of the 1,192 medical errors, 183 (15.4%) in 128 (9.3%) patients were adverse events that were followed by one or more clinical consequences (n = 163) or that required one or more procedures or treatments (n = 58). By multivariable analysis, having two or more adverse events was an independent risk factor for ICU mortality (odds ratio, 3.09; 95% confidence interval, 1.30–7.36; P = 0.039).
Conclusions: The impact of medical errors on mortality indicates an urgent need to develop prevention programs. We have planned a study to assess a program based on our results.
In intensive care units (ICUs), the complexity of care and severity of illnesses result in a high risk for iatrogenic events. Medical errors are common and cause morbidity and mortality in critically ill patients.
After careful adjustment for severity of illness, experiencing more than two adverse events was associated with a threefold increase in the risk of ICU death.
Critically ill patients are highly vulnerable to medical errors, because they usually have both underlying comorbidities and acute organ dysfunctions (8). In addition, the life-sustaining treatments and highly technical routine care used in ICUs provide many opportunities for medical errors. Efforts have been made to develop standardized definitions of medical errors (9, 10). To be good-quality indicators, medical errors must be common, preventable, reproducible, easy to diagnose and to collect, associated with high morbidity and mortality, and easy to report without fear of punishment. In critical care, there is no generally accepted list of medical errors exhibiting these characteristics.
The objectives of this study were to select medical errors suitable for use as quality indicators, using a modified Delphi technique (part I of the IATROREF I Study) and then to evaluate the incidence of these medical errors in French ICUs (part II) and to assess their relationship with mortality. Some of the results of this study have been previously reported in the form of an abstract (11).
See the online supplement for details of the Delphi process used to select the indicators. We defined a medical error as the failure of a planned action to be completed as intended (i.e., error of execution) or the use of a wrong plan to achieve an aim (i.e., error of planning) and an adverse event as an injury caused by a medical intervention that resulted in harm (12).
An invitation to participate in the study was sent by mail in January 2006 to the directors of 250 ICUs having more than 6 beds, and who belonged to the French Society for Critical Care Medicine. Seventy closed ICUs in public and private institutions agreed to participate in the study.
The study period was from March 27, 2006, at 8 a.m., to April 3, 2006, at 8 a.m. No holidays occurred during this period, making full staffing likely. We included all patients who were in study ICUs on at least 1 day of the study week. Readmitted patients were also included. The study was approved by the Advisory Committee for the Protection of Individuals involved in Biomedical Research, which waived the requirement for written informed consent.
The following ICU characteristics were recorded: hospital (university, community, or private hospital; number of beds; and whether the hospital had a risk-management unit), structure of the unit (type, number of acute and intermediate beds, number of senior physicians and fellows, physician-to-patient ratio, and nurse-to-patient ratio), shift for nurses and assistant nurses, whether the unit had a patient-safety program, whether deaths were routinely discussed during specific meetings, ICU mortality, and ICU length of stay. We recorded whether the unit had written procedures for weaning off mechanical ventilation, dialysis, sedation, insulin, and anticoagulant prescription; as well as the extent to which these guidelines were known and followed by the ICU staff.
In each center, investigators were free to choose between paper and an electronic file at a data collection website (http://www.iatroref.outcomerea.fr/iatroref/). The same data were collected by both methods. Data written on paper were later entered into the computer, using software that automatically detected inconsistencies. At the end of the study, missing data and inconsistencies were resolved by E-mails sent to each center. Date of hospital discharge was censored on June 30, 2006. The following characteristics were recorded: demographic characteristics (age, sex); underlying diseases, using the Knaus classification (13); admission category (medical, scheduled surgery, or unscheduled surgery), invasive procedures (number of arterial or venous central lines, number of peripheral intravenous devices, number of days of mechanical and noninvasive ventilation), reason for admission (with nine categories defined prospectively before the study, namely, respiratory, cardiac, or renal failure; coma; multiple organ failure, acute exacerbation of chronic pulmonary disease, monitoring, trauma; and scheduled surgery), and number of days with specific medication categories (preventive and curative anticoagulants, sedatives, neuromuscular blockers, insulin, vasoactive drugs, and inotropic drugs). The Simplified Acute Physiology Score (SAPS II) (14) at admission was computed using the worst physical and laboratory data during the first 24 hours in the ICU. Stay duration in the ICU and acute care hospital and vital status at ICU and hospital discharge were recorded.
The study steering committee wrote a guide for the investigators, explaining how to collect data on patients and medical errors. The guide supplied the definitions (Table 1) and a list of common situations in which each medical error might occur. The ICU physician in charge of the study in each center, assisted by a head nurse, evaluated the consequences of each medical error using preestablished lists and assessed severity on a 6-grade scale (1, no change in management; 2, clinical monitoring; 3, additional laboratory or radiological investigations; 4, additional medical or surgical treatment; 5, initiation of treatment for organ dysfunction; and 6, contribution to death). Contribution to death was assessed as none, improbable, not very probable, somewhat probable, probable, or certain. For 1 week, a 24-hour hotline (M.G.O. or L.S.) was open for answering queries about the study. A daily E-mail was sent to each center to provide support.
Medical Error | Definition |
|---|---|
| Suction circuit failure during intubation | The suction system does not work properly: The pressure decrease is not sufficient to ensure removal of pharyngeal, gastric, and/or bronchial secretions during intubation |
| Laryngoscope dysfunction | The laryngoscope does not work properly: The light is not strong enough or does not turn on during laryngoscopy, assembly of the blades on the handle is difficult or impossible, there is no contact |
| Medication administered to wrong patient | Medication intended for patient A is given to patient B |
| Error administering anticoagulant medication | Anticoagulant therapy is not given as prescribed. The divergence may relate to the planning and/or execution of the prescription: drug given, dosage, preparation and administration modalities, dosing times, or dosing intervals |
| Error prescribing anticoagulant medication | Failure to comply with recommendations (learned societies, department protocols, local drug committees) regarding the indications, dosage, administration modalities, contraindications, drug interactions, or laboratory monitoring of anticoagulant treatment |
| Error administering vasoactive drugs | Vasoactive therapy is not given as prescribed. The divergence may relate to the planning and/or execution of the prescription: drug given, dosage, or preparation and administration modalities |
| Error administering insulin | Insulin therapy is not given as prescribed (including as per department protocol). The divergence may relate to the planning and/or execution of the prescription: drug given, dosage, or preparation and administration modalities |
| Accidental removal of a central venous catheter | Unplanned complete removal of a central venous catheter by the patient or by a health care worker during care or manipulation of the catheter |
| Accidental extubation | Unplanned extubation |
| Failure to place patient in semirecumbent position, in the absence of contraindication, during invasive mechanical ventilation with enteral nutrition | A patient receiving enteral nutrition is not kept in a 30–45 degree semirecumbent position during invasive ventilation.For this indicator, patients were excluded if they had a contraindication to the semirecumbent position (hemodynamic instability, spinal surgery, injury to the thoracolumbar spine, or unstable fracture of the pelvis) or if they were in the prone position |
| Overinflation of the endotracheal balloon | Mean pressure in the endotracheal balloon, measured with a manometer and recorded on the medical chart, is equal to or greater than 35 cm H2O |
| Pneumothorax related to insertion of a central venous catheter | Partial or complete pleural detachment by a gaseous effusion on the same side as insertion (or attempted insertion) of a catheter into the internal jugular or subclavian vein, occurring within 48 h of insertion (or attempted insertion), diagnosed radiologically or diagnosed clinically, with a need for drainage of such urgency as to preclude previous radiography |
| Fall | The patient falls |
| Delay in surgical treatment | Excessive time between the diagnosis of an acute condition requiring surgery and the surgical procedure according to good clinical practice. Surgery must be performed with no delay at all in patients who have immediately life-threatening lesions (e.g., rupture of large vessels, aortic dissection, or ectopic pregnancy). Surgery must be performed within 6 h of the diagnosis of other lesions (e.g., compound fracture, peritonitis, or acute limb ischemia) |
Of the 70 centers, 16 (22.8%) were selected at random for an audit during the study. For practical reasons, two centers were audited on Day 2, one on Day 3, three on Day 4, two on Day 5, and eight on Day 8. The audits were conducted by eight intensivists, each with at least 2 years of ICU experience, who were trained by three study investigators (J.F.T., M.G.O., and L.S.). The auditors were given a guide that contained methods for collecting data on patients and medical errors; definitions of selected medical errors and their consequences; and grades for reporting severity, preventability, and relation with death.
Patient and center variables are reported as numbers (percentage) for qualitative data and medians (Q1–Q3) for quantitative data. Numbers of medical errors on each day were compared by Cochran-Armitage test. Reproducibility of the clinical and laboratory data and of the occurrence of medical errors was tested using κ coefficients or intraclass correlations. Center-based risk factors of medical errors were determined by comparing incidences according to the characteristics of the units, using the Wilcoxon test or Kruskal-Wallis test, as appropriate. We used conditional logistic regression to identify patient-based risk factors for medical errors, taking into account stratification on center and the time spent in the study. Parameters significantly associated with the risk of medical errors were introduced in a hierarchical multivariate logistic mixed model using the GLIMMIX procedure available in SAS 9.13 (SAS Institute, Cary, NC). A backward selection procedure was used to select independent risk factors with P < 0.05, and results were routinely adjusted on time spent in the study.
To estimate the relationship between medical errors and ICU death, we built a multivariate model using conditional logistic regression with stratification by center and the time spent in the study. A stepwise process (SLE = 0.2, SLS = 0.05) was used to select a subset of predictive variables from this initial list of variables of clinical interest: age, severity of illness at admission (SAPS II) (14), comorbidity, symptoms and diagnosis at ICU admission, transfer from ward, mechanical ventilation, central venous and arterial catheters, vasoactive drugs, anticoagulants, insulin, sedation, and neuromuscular blockers. The number of medical errors and the number of adverse events were then tested separately in a multivariate hierarchical logistic mixed model. All hierarchical logistic mixed models included random intercept to reflect center variability and were adjusted on time spent in the study. All tests were two-sided, and statistical analyses were performed with SAS 9.13.
The 70 study ICUs included 1,377 patients. After exclusion of the 8 patients who were entered twice, 1,369 patients were left for the study. ICU and patient characteristics are shown in Tables 2 and 3, respectively. Almost half the ICUs had written procedures for respiratory care or insulin or anticoagulant treatment, which were followed in most ICUs.
Variable | Data | |
|---|---|---|
| Hospital | ||
| University hospital, n (%) | 35 (50) | |
| Number of hospital beds, median (25th percentile, 75th percentile) | 642 (455–1,000) | |
| Presence of a risk management unit, n (%) | 55 (78.5) | |
| ICU | ||
| Medical, n (%) | 21 (30) | |
| Surgical, n, (%) | 7 (10) | |
| Mixed, n (%) | 42 (52.9) | |
| Number of acute beds per unit, mean ± SD | 13.1 ± 4.8 | |
| Number of intermediate beds per unit, mean ± SD | 2.7 ± 4.2 | |
| Number of units without intermediate beds, n (%) | 42 (60) | |
| Number of units with a safety-reporting program, n (%) | 37 (52.8) | |
| Meetings on ICU deaths, n (%) | 23 (32.8) | |
| Number of attending physicians, mean ± SD | 5.46 ± 1.91 | |
| Number of junior physicians, mean ± SD | 2.75 ± 2.06 | |
| Day off after duty for physicians, n (%) | 61 (87.1) | |
| Training of junior physicians on duty, n (%) | 40 (57.1) | |
| Patient-to-physician ratio | ||
| Mean ± SD | 3.05 ± 1.12 | |
| Median (25th percentile, 75th percentile) | 2.78 (2.23–3.72) | |
| Patient-to-nurse ratio, day (night) | ||
| Mean ± SD | 2.72 ± 0.45 (3.14 ± 0.68) | |
| Median (25th percentile, 75th percentile) | 2.6 (2.5–3) | |
| Patient-to-nursing assistant ratio, day (night), mean ± SD | 4.26 ± 1.54 (5.35 ± 3.09) | |
| 12-h shifts for nurses | 36 (51.4) | |
| Years of ICU experience for nurses, mean ± SD | 4.09 ± 2.05 | |
| Written procedures* | ||
| Weaning off mechanical ventilation, n (%) | ||
| None | 37 (52.8) | |
| Known and almost always or always followed | 21 (30) | |
| Unknown or not followed | 12 (17.1) | |
| Sedation, n (%) | ||
| None | 40 (57.1) | |
| Known and almost always or always followed | 21 (30) | |
| Unknown or not followed | 9 (12.8) | |
| Insulin treatment, n (%) | ||
| None | 29 (41.4) | |
| Known and almost always or always followed | 34 (48.5) | |
| Unknown or not followed | 7 (10) | |
| Anticoagulant, n (%) | ||
| None | 48 (68.5) | |
| Known and almost always or always followed | 12 (17.1) | |
| Unknown or not followed | 10 (7) | |
Variable | Data | |
|---|---|---|
| Age, years, mean ± SD | 61.2 ± 17.8 | |
| Male sex, n (%) | 896 (65.4) | |
| SAPS II at admission, mean ± SD | 45.4 ± 19.8 | |
| Admission type, n (%) | ||
| Medical | 964 (70.4) | |
| Scheduled surgery | 136 (9.9) | |
| Unscheduled surgery | 254 (18.5) | |
| Reason for admission, n (%) | ||
| Respiratory failure | 382 (27.9) | |
| Cardiovascular failure | 330 (24.1) | |
| Renal failure | 58 (4.3) | |
| Coma | 212 (15.4) | |
| Multiorgan failure | 50 (3.6) | |
| Acute exacerbation of COPD | 39 (2.8) | |
| Trauma | 52 (3.7) | |
| Monitoring | 151 (11.0) | |
| Scheduled surgery | 64 (4.6) | |
| Comorbid conditions, n (%) | 857 (62.6) | |
| Procedures ≥1 day, n (%) | ||
| Mechanical ventilation | 805 (58.8) | |
| Noninvasive ventilation | 171 (12.4) | |
| Central venous catheter | 794 (57.9) | |
| Treatments ≥1 day, n (%) | ||
| Insulin | 801 (58.5) | |
| Vasoactive drugs | 428 (31.2) | |
| Prophylactic anticoagulation | 812 (59.3) | |
| Curative anticoagulation | 247 (18.0) | |
| Sedatives | 565 (41.1) | |
| Length of ICU stay, days | ||
| Mean ± SD | 22.3 ± 33.5 | |
| Median (25th percentile, 75th percentile) | 10 (4–26) | |
| Mortality, n (%) | ||
| ICU | 285 (20.8) | |
| Hospital | 350 (25.5) | |
During the study, 1,192 medical errors were reported for 1,369 patients. At least one medical error occurred in 367 (26.8%) patients, for a rate of 2.1/1,000 patient-days. The rate of second occurrences of specific errors was 203/367 (55.3%). The most common medical error was error in insulin administration, with a frequency of 185.9/1,000 days of insulin treatment. Insulin was used in 801 (801/1,369, 58.5%) patients, according to an individually tailored protocol in 30 centers and to a standardized protocol in 40 centers. We examined the characteristics of 36 (90%) of the 40 standardized protocols and found no correlation with insulin administration errors in a hierarchical negative binomial regression model (data not shown). Table 4 reports the frequency of medical errors, according to the number of days in the study and to the procedure or treatment related to the medical error. We found no relation between the hour of reporting (occurrence or discovery of the medical error) and the occurrence of each medical error (data not shown). Table E2 (see the online supplement) displays the results of the univariate analysis of risk factors for medical errors involving selected ICU and patient characteristics. Only patient-related factors were significant and were entered into the multivariable model. In the multivariable model, factors independently associated with having at least one medical error (odds ratio [OR], 95% confidence interval [95% CI]; P value) included scheduled surgery (0.54, 0.31–0.96; P = 0.035), insulin prescription during the study week (1.60, 1.10–2.35; P = 0.015), central venous catheter during the study week (1.80, 1.22–2.64; P = 0.003), mechanical ventilation during the study week (2.41, 1.64–3.55, P ≤ 0.0001), and number of days in the study (1.46, 1.35–1.58; P < 0.0001). The intercenter covariance parameter estimate was 1.72 (SD, 0.40), indicating highly significant residual variability across centers that was not explained by the model. The independent risk factors for experiencing at least one medical error could be used to identify groups of patients that differ regarding their levels of risk. Table E3 (see the online supplement) displays the observed risk of medical errors according to the number of criteria met (patients admitted for emergency surgery, use of mechanical ventilation, central venous catheter, or insulin prescription during follow up). For example, patients meeting at least three of these four criteria (i.e., 55.3% of the overall population) contributed 65.9% of all medical errors.
Medical Error | Number of Medical Errors | Domain of Care with Opportunity for Medical Errors* | Patients with Domain of Care | Days with Domain of Care† | ME/1,000 Days with Domain of Care; Median (IQR) | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| Suction circuit failure during intubation | 2 | IMV | 805 | 3,223 | 0.6; 0 (0) | |||||
| Laryngoscope dysfunction | 0 | IMV | 805 | 3,223 | 0; 0 (0) | |||||
| Medication administered to wrong patient | 23 | All | 1,362 | 5,678 | 4.1; 0 (0) | |||||
| Error administering anticoagulant medication | 23 | AC | 1,033 | 4,362 | 5.3; 0 (9) | |||||
| Error prescribing anticoagulant medication | 36 | AC | 1,033 | 4,362 | 8.3; 0 (9) | |||||
| Error administering vasoactive drugs | 29 | VAD | 428 | 1,379 | 21.0; 0 (0) | |||||
| Error administering insulin‡ | 630 | Insulin | 801 | 3,389 | 185.9; 0 (34) | |||||
| Accidental removal of a central venous catheter | 9 | CVC | 794 | 3,437 | 2.6; 0 (0) | |||||
| Accidental extubation | ||||||||||
| Accidental extubation | 14 | IMV | 805 | 3,223 | 4.3; 0 (0) | |||||
| Self-extubation | 21 | IMV | 805 | 3,223 | 6.5; 0 (0) | |||||
| Failure to place patient in semirecumbent position, in the absence of contraindication, during invasive artificial ventilation with enteral nutrition | 121 | IMV | 805 | 3,223 | 37.5; 0 (48) | |||||
| Overinflation of intubation catheter balloon | 261 | IMV | 805 | 3,223 | 81.0; 0 (73) | |||||
| Pneumothorax related to insertion of central venous catheter | 7 | CVC | 794 | 3,437 | 2.0; 0 (0) | |||||
| Fall | 6 | All | 1,369 | 5,678 | 1.1; 0 (0) | |||||
| Delay in surgical treatment | 10 | All | 1,369 | 5,678 | 1.8; 0 (0) | |||||
Of the 1,192 medical errors, 183 (15.4%) in 128 (9.3%) patients were classified as adverse events and were followed by one or more clinical consequences (n = 163) or required one or more procedures or treatments (n = 58). Clinical monitoring was required for 74/1,192 (6.2%) events, laboratory or radiological investigations for 28/1,192 (2.3%) events, medical or surgical treatment for 30/1,192 (2.5%) events, and treatment for organ dysfunction for 2/1,190 events. Four deaths were attributable to adverse events (one after self-extubation and three after delayed surgical treatment). One cardiac arrest occurred after accidental extubation; there were no neurological sequelae and the patient died 2 months later from multiple organ failure induced by nosocomial infection. Severity was greater than level 3 (need for monitoring or radiological investigation) for the following numbers of events: accidental extubation (n = 4), overinflation of the intubation catheter balloon (n = 2), failure to place the patient in the semirecumbent position in the absence of contraindications (n = 2), error administering vasoactive drugs (n = 5), error administering insulin (n = 15), error prescribing anticoagulation medication (n = 2), pneumothorax related to insertion of a central venous catheter (n = 2), delay in surgical treatment (n = 7), and error administering anticoagulation medication (n = 1). For 44/1,192 (3.7%) medical errors, no data were obtained about possible consequences. No consequences were recorded for suction circuit failure during intubation, falls, or medications given to the wrong patient. Table E4 (see the online supplement) displays the consequences, including additional treatments, of the other medical errors.
The multivariable conditional logistic regression model identified 11 variables independently associated with death (see Table E5 in the online supplement). After adjustment for these variables in the hierarchical model, we found no significant association between the number of medical errors or the occurrence of the most frequent medical error and ICU death (Table 5). However, having more than two adverse events increased the risk of death (OR, 3.09; 95% CI, 1.30–7.36). As shown in Table E6 (see the online supplement), this association persisted when the analysis was restricted to mechanically ventilated patients (OR, 3.08; 95% CI, 1.29–7.36). Clustering of events had no effect on the prognostic model.
Survived | Died | Adjusted OR* (95% CI) | P Value | |
|---|---|---|---|---|
| Number of medical errors | 0.65 | |||
| 0 | 772 | 230 | 1 | |
| 1 | 117 | 47 | 0.88 (0.56–1.40) | |
| 2 | 37 | 20 | 1.04 (0.52–2.11) | |
| 3 | 28 | 16 | 1.73 (0.79–3.78) | |
| 4 | 32 | 12 | 0.81 (0.37–1.77) | |
| ≥5 | 33 | 25 | 1.29 (0.67–2.49) | |
| Number of medical errors followed by clinical or therapeutic consequences = adverse events | 0.039 | |||
| 0 | 940 | 301 | 1 | |
| 1 | 63 | 30 | 1.09 (0.64–1.88) | |
| ≥2 | 16 | 19 | 3.09 (1.30–7.36) | |
| Number of medical errors without consequences | 0.57 | |||
| 0 | 806 | 247 | 1 | |
| 1 | 98 | 42 | 0.81 (0.50–1.32) | |
| 2 | 38 | 17 | 0.89 (0.44–1.81) | |
| 3 | 25 | 14 | 1.72 (0.78–3.79) | |
| ≥4 | 52 | 30 | 0.95 (0.53–1.69) | |
| At least one “accidental extubation”† | 22 | 9 | 0.98 (0.39–2.46) | 0.93 |
| At least one “overinflation of the intubation catheter balloon”† | 71 | 42 | 1.01 (0.61–1.67) | 0.98 |
| At least one “failure to place patient in semirecumbent position”† | 58 | 27 | 0.96 (0.52–1.77) | 0.89 |
| At least one “error administering insulin”‡ | 95 | 52 | 1.27 (0.80–2.02) | 0.31 |
| At least one “error prescribing anticoagulant medication”§ | 19 | 8 | 0.73 (0.26–2.03) | 0.55 |
Audits were performed in 16 of 70 (22.8%) ICUs, for a total of 193 patients and 576 ICU days. Agreement was perfect for accidental removal of central venous catheter, falls, and extubation (Kappa, 95% CI, K = 1). Agreement was substantial for failure to place in the semirecumbent position with mechanical ventilation and enteral nutrition (K = 0.74; 95% CI, 0.4–1), pneumothorax (K = 0.66; 95% CI, 0.05–1.0), overinflation of the intubation catheter balloon (K = 0.78; 95% CI, 0.64–0.92), insulin administration errors (K = 0.74; 95% CI, 0.64–0.84), anticoagulant administration errors (K = 0.72; 95% CI, 0.42–1), and anticoagulant prescription errors (K = 0.70; 95% CI, 0.43–0.98). Agreement was almost perfect for medication given to the wrong patient (K = 0.89; 95% CI, 0.75–1). For errors in administering vasoactive drugs, agreement was moderate (K = 0.43. 95% CI, 0.17–0.69).
We studied the incidence in 70 French ICUs of medical errors selected using Delphi techniques by a large group of professionals belonging to different specialties involved in critical care. The medical error rate was 2.1/1,000 patient-days. The most common medical errors were insulin administration errors. Of all medical errors, we found 15.4% adverse events. Adverse events (i.e., medical errors with either therapeutic or clinical consequences) had considerable prognostic significance, with a threefold increase in mortality among patients who experienced more than two such events.
Among our patients, 26.8% experienced at least 1 of the 14 selected medical errors. In earlier studies, the incidence of all medical errors or adverse events in ICU patients ranged from 6.9 to 56.2% (15–17). Our results are consistent with data reported by De Lassence and colleagues (18) regarding self-extubation (6.4/1,000 vs. 6.5/1,000 in our study) and accidental extubation (3.0/1,000 vs. 4.3/1,000 in our study). The incidence of accidental removal of central venous catheters (2.6/1,000) was in line with data by Lorente and colleagues (2.02/1,000) (19). The 0.88% rate of pneumothorax related to central venous catheters in our patients was lower than usually reported (20, 21). Underreporting is a major obstacle to studies of medical errors. Reasons that may lead to underreporting by staff members include lack of clear definitions of errors (22), fear of punishment, insufficient emphasis on patient safety in the unit leading to inadequate motivation of the staff (23–25), and absence of feedback about the effects of medical errors (22, 26). We sought to avoid these limitations. Thus, we used medical errors that met criteria for good-quality indicators, supplied a detailed definition of each error, and prepared an investigator's guide on data collection. Nevertheless, differences occurred across centers, suggesting underreporting, although we used the reporting method recommended for epidemiological studies (27). The reproducibility of medical error event reporting was good, except for vasoactive drug administration errors. The physicians who performed the audits believed that chart review was usually inadequate for detecting medical errors. Data collection by an external team of ICU staff members (28–30) was strongly recommended by the steering committee but proved infeasible. Collecting all medical errors all the time would be an overwhelming task. The results of this study have helped the French Society of Critical Care Medicine to select medical errors for the purpose of mandatory reporting in French ICUs.
The patients with the most severe illness were more often the subject of medical errors. Only patient variables were selected by the multivariable model, in line with several earlier studies (17, 20, 31, 32). Prevention programs should target the most severely ill patients, defined by our multivariable model as patients receiving mechanical ventilation, central venous catheterization, or insulin or not undergoing scheduled surgery. The considerable center-to-center variability in medical error rates might be partly ascribable to the short study period but also suggests a potential for improvement. Many system factors that might explain the variability (33) were not investigated and deserve further studies. More specifically, workload, burnout (34), and teamwork climate or job satisfaction (33) may affect medical error rates (35). This multiplicity of factors considerably complicates the development of a safety culture in the ICU, and studies of sources of variability across centers will likely be helpful in the future. Furthermore, our data were obtained in ICUs whose patient-to-nurse ratios were those usually encountered in France. We found no clustering of adverse events at specific times during the 24-hour cycle or on weekends (36, 37). The relation between the complex ICU environment, complexity of the patients, and occurrence of medical errors suggests that guidelines, education, and communication skills training (38) might foster a culture of safety in the ICU, thereby decreasing medical error rates.
The five most common adverse events were errors administering medications (vasoactive drugs and insulin) and events related to mechanical ventilation (unplanned extubation, overinflation of intubation catheter balloon, and failure to place the patient in the semirecumbent position). These results are consistent with those of the Sentinel Events Evaluation (SEE) Study (39). Insulin administration errors deserve further comment. This adverse event was selected by the experts after the first randomized study of tight glucose control (40). Our definition, which included incidents without harm, contributes to explain the high frequency of this event. Given the premature discontinuation of two studies on tight glucose control (41) because of increased rates of hypoglycemia, the results of a meta-analysis (42), and the potential harm associated with hypoglycemia (43), the risk-to-benefit ratio of tight glucose control is still controversial in ICUs. Although insulin administration errors did not consistently cause harm in our study, the potential for harm existed. Therefore, prevention programs must focus on this medical error, as done in part III of this research project, whose results have not yet been published. Had the high rate of insulin administration errors been known at the time, the risk-to-benefit ratio of studies of tight glucose control would perhaps have been better assessed before the conduct of large randomized studies that exposed patients to harm.
Having more than two adverse events was an independent risk factor for ICU mortality. This result agrees with findings from other studies that used different methodologies (20, 43, 44). The mortality attributable to adverse events is difficult to assess for many reasons including case mix, confounding factors for mortality, and occurrence of multiple adverse events in the same patient (45, 46) with interactions between adverse events. When these factors were taken into account using appropriate statistical methods, we found that having multiple adverse events correlated with death, in keeping with an earlier study (8).
Strengths of this study include the selection of medical errors by a panel of experts, clear definitions of errors, the audit of the study, and the careful assessment of effects on mortality. These results and the preventability analysis (data not shown) permitted us to design a multifaceted prevention program, which will be evaluated in part III of our IATROREF research project. This study has several limitations. First, the 1-week study period may be too short to allow generalization of our results. Second, considerable reporting differences occurred across centers. Some medical errors were perhaps not reported, despite the brief study duration and daily supportive E-mail to each ICU team. Although we supplied detailed definitions of the medical errors, there may have been some measure of personal interpretation. Differences in motivation and in the emphasis on patient safety, together with the lack of in-person training about the study protocol, probably contributed to the differences in reporting and in the interpretation of clinical consequences, despite the existence of a collection guide. Third, few explanations to the medical errors were suggested. More specifically, the impact of the lack of guidelines in about half the centers and of failure to follow guidelines in a few centers cannot be determined without a root cause analysis of the medical errors.
In conclusion, our study provides new insights into the epidemiology of selected medical errors in the ICU. We established a list of medical errors that can serve as quality indicators in ICUs and we estimated their incidence. The impact of adverse events on mortality indicates an urgent need for prevention programs in ICUs. We have planned a study to evaluate such a program.
The authors thank A. Wolfe, M.D., for assistance in preparing the manuscript. The authors are grateful to the French Clinical Research Committee (PHRC) and the National Authority for Health for supporting this study. The authors are indebted to the staff members of the 70 study centers for their help. The authors thank François Philippart, M.D., Vincent Willems, M.D., Jean Ralph Zahar, M.D., and Cathy Duc, M.D., for their participation in the audit of the study.
OUTCOMEREA Study Group: Scientific Committee: Jean-François Timsit, M.D., Professor, ICU, Epidemiologie Pronostique des Affections Graves, Institut Albert Bonniot, INSERM, UJF, U823, La Tronche and Medical ICU, Hôpital Albert Michallon, Grenobel; Gilles Troché, M.D., Staff Physician, Hôpital Mignot, Le Chesnay; Adrien Descorps-Declere, Staff Physician, Hôpital Antoine, Béclère, Clamart; Pierre Moine, M.D., Ph.D., Staff Physician, DAR, Hôpital Lariboisière, Paris; Elie Azoulay, M.D., Professor, ICU, Hôpital Saint Louis, Paris; Yves Cohen, M.D., Professor, ICU, Hôpital Avicenne, Bobigny; Maité Garrouste-Orgeas, M.D., Staff Physician, ICU, Groupe Hospitalier Paris Saint Joseph, Paris, France, Epidemiologie Pronostique des Affections Graves, Institut Albert Bonniot, INSERM, UJF, U823, La Tronche; Jean-Philippe Fosse, M.D., Staff Physician, ICU, Hôpital Avicenne, Bobigny; Lilia Soufir, M.D., Staff Physician, DAR, Groupe Hospitalier Paris Saint Joseph, Paris; Jean-Ralph Zahar, M.D., Attending Physician, Microbiology Department, Hôpital Necker, Paris; Christophe Adrie, M.D., Staff Physician, ICU, Hôpital Delafontaine, Saint Denis; Jean Carlet, M.D., Quality and Safety Consultant, Groupe Hospitalier Paris Saint Joseph, Paris; Alexandre Lautrette, M.D., Staff Physician, ICU, Hôpital Gabriel Montpied, Clermont Ferrand; A. Le Monnier, M.D., Attending Physician, Microbiology Department, Hôpital Général, Versailles; M. Darmon, M.D., Staff Physician, ICU, Hôpital Nord, Saint Etienne; D. Golgran-Toledano, M.D., Staff Physician, ICU, Hôpital Général, Gonesse; C. Clec'h, M.D., Staff Physician, ICU, Hôpital Avicenne, Bobigny; B. Souweine, M.D., Professor, ICU, Hôpital Gabriel Montpied, Clermont Ferrand, France.
Biostatistical and informatics expertise: Sylvie Chevret, M.D., Ph.D., Professor, Medical Computer Sciences and Biostatistics Department, Hôpital Saint Louis, Paris; Corinne Alberti, M.D., Staff Physician, Medical Computer Sciences and Biostatistics Department, Robert Debré, Paris; Frederik Lecorre, Informatics Engineer, Supelec; Didier Nakache, Informatics Engineer, Conservatoire National des Arts et Métiers (CNAM), Paris; A. Vesin, Biostatistician, OUTCOMEREA Department, Grenoble; A. Français, Biostatistician, OUTCOMEREA Department, Grenoble, France; Clinical research organization monitors: C. Tournegros, N. Kaddour, L. Ferrand, B. Berthe.
Participating centers: P. Arnodo, M.D., P. Maurin, R.N., Medical–Surgical ICU, Centre Hospitalier, Aix en Provence; T. Finge, M.D., Medical–Surgical ICU, Ales; C. Santré, M.D., M. Vinclair, M.D., Medical–Surgical ICU, Hôpital Régional, Annecy; V. Cadiergues, M.D., Medical–Surgical ICU, Hôpital Général, Annonay; G. Rigal, M.D., A. Dessenne, R.N., Medical–Surgical ICU, Hôpital Général, Arles; C. Clec'h, M.D., C. Vallon, R.N., Medical–Surgical ICU, Hôpital Avicenne, Bobigny; E. L'Her, M.D., Medical ICU, Centre Hospitalier Universitaire de la Cavale Blanche, Brest; F. Berkelmans, M.D., H. Peneau, M.D., Medical–Surgical ICU, Hôpital Général, Calais; J. M. Korach, M.D., C. Pionnier, R.N., Medical–Surgical ICU, Centre Hospitalier Général, Chalons en Champagne; O. Gontier, M.D., M. Perdereau, R.N., Medical–Surgical ICU, Hôpital Louis Pasteur, Chartres; B. Souweine, M.D., F. Thiollière, M.D., Renal and Medical ICU, Hôpital Gabriel Montpied, Clermont Ferrand; J. E. Bazin, M.D., J. M. Constantin, M.D., Adult ICU, Hôpital Hotel Dieu, Clermont Ferrand; G. Laplatte, M.D., M. L. Ueber, R.N., Medical ICU, Hôpitaux Civils, Colmar; G. Chevrel, M.D., A. De Lassence, M.D., Medical ICU, Hôpital Louis Mourier, Colombes; M. Darmon, M.D., I. Bourgeon, R.N., Medical ICU, Hôpital Henri Mondor, Créteil; J. P. Rigaud, M.D., I. Auriant, M.D., Medical–Surgical ICU, Hôpital Général, Dieppe; J. P. Colin, M.D., S. Jamali, M.D., ICU, Centre Hospitalier Dourdan, Dourdan; A. Conia, M.D., A. Guillon, R.N., Medical–Surgical ICU, Centre Hospitalier Général, Dreux; M. Kaidomar, M.D., V. Filippi, R.N., ICU, Hôpital Général, Frejus; V. Maxime, M.D., D. Annane, M.D., Medical ICU, Hôpital Raymond Poincaré, Garches; D. Golgran-Toledano, M.D., Medical–Surgical ICU, Hôpital Général, Gonesse; J. F. Timsit, M.D., C. Schwebel, M.D., Medical ICU, Hôpital Albert Michallon, Grenoble; M. Durand, M.D., F. Gautard, R.N., Thoracic and Cardiovascular ICU, Hôpital Albert Michallon, Grenoble; C. Jacquot, M.D., M. C. Herault, R.N., Surgical ICU, Hôpital Albert Michallon, Grenoble; C. Cheval, M.D., Medical–Surgical ICU, Hôpital Marie José Treffot, Hyères; J. Reignier, M.D., Y. Alcourt, R.N., Medical Surgical ICU, Centre Hospitalier Départemental les Oudaries, La Roche Sur Yon; G. Troché, M.D., S. Legriel, M.D., Medical–Surgical ICU, Le Chesnay; F. Saulnier, M.D., S. Soubrier, M.D., Medical ICU, Hôpital Albert Calmette, Lille; L. Montesino, M.D., Medical-Surgical ICU, Hôpital Général, Longjumeau; A. Caubel, M.D., Medical–Surgical ICU, Centre Hospitalier Bretagne Sud, Lorient; M. Bret, M.D., E. Croizat, R.N., Medical–Surgical ICU, Centre Hospitalier Universitaire, Hospices Civils, Lyon; B. Allaouchiche, M.D., S. Pereira, R.N., Medical–Surgical ICU, Hôpital Edouard Herriot, Lyon; I. Mohammedi, M.D., Medical ICU, Hôpital Edouard Herriot, Lyon; B. Floccard, M.D., S. Pereira, R.N., ICU, Hôpital Edouard Herriot, Lyon; N. Embriaco, M.D., L. Papazian, M.D., Medical ICU, Hôpital Sainte Marguerite, Marseille; J. F. Poussel, M.D., P. Cangemi, M.D., Medical–Surgical ICU, Hôpital Notre Dame Du Bon Secours, Metz; F. Moliere, M.D., O. Jonquet, M.D., Medical ICU, Centre Hospitalier Universitaire, Gui de Chauliac, Montpellier; K. Kuteifan, M.D., C. Meyer, M.D., Medical ICU, Hôpital Emile Muller, Mulhouse; A. Croizy, M.D., P. E. Bollaert, M.D., Medical ICU, Hôpital Central, Nancy; C. Bretonnière, M.D., D. Villers, M.D., Medical ICU, Centre Hospitalier Universitaire Hotel Dieu, Nantes; J. Y. Lefrant, M.D., F. Casano, R.N., Surgical ICU, Anesthesiology, Pain, and Emergency Department, Centre Hospitalier Universitaire Caremeau, Nîmes; J. M. Mazou, M.D., Medical–Surgical ICU, Hôpital Général, Oloron Sainte Marie; Y. Monseau, M.D., P. Delour, M.D., Medical ICU, Hôpital Général, Périgueux; A. Rabbat, M.D., A. Lefevre, M.D., Medical ICU, Hôpital Hôtel Dieu, Paris 1; F. Fieux, M.D., E. Leroux, R.N., Surgical ICU, Hôpital Saint Louis, Paris 10; I. Coquet, M.D., M. Adda, M.D., Medical ICU, Hôpital Saint Louis, Paris 10; M. Garrouste-Orgeas, M.D., F. Diaw, R.N., Medical–Surgical ICU, Groupe Hospitalier Paris Saint Joseph, Paris 14; L. Soufir, M.D., N. Kaddour, CRO, Department of Anesthesiology, Groupe Hospitalier Paris Saint Joseph, Paris 14; J. Charpentier, M.D., L. Laronche, R.N., Medical ICU, Groupe Hospitalier Cochin, Paris 14; C. Faisy, M.D., J. Y. Fagon, M.D., Medical ICU, Hôpital Européen Georges Pompidou, Paris 15; F. Bellenfant, M.D., M. C. Soyer, R.N., Traumatologic and Surgical ICU, Hôpital Européen Georges Pompidou, Paris 15; B. Mourvillier, M.D., B. Reignier, M.D., Medical ICU, Hôpital Bichat, Paris 18; C. Ridel, M.D., M. Neveur, R.N., Nephrology and Transplant ICU, Hôpital Tenon, Paris 20; M. Closon, M.D., A. Lepape, M.D., Surgical ICU, Hôpital Lyon Sud, Pierre Bénite, Lyon; J. Bailly Salin, M.D., J. C. Lacherade, M.D., Medical–Surgical ICU, Poissy; M. Pinsart, M.D., P. Schall, R.N., Medical ICU, Hôpital Jean Bernard, Poitiers; E. Boulet, M.D., A. Torelli, R.N., Medical–Surgical ICU, Hôpital René Dubos, Pontoise; A. Gacouin, M.D., B. Gaillard, R.N., Infectious and Medical ICU, Hôpital Pontchaillou, Rennes; F. Tamion, M.D., Medical ICU, Centre Hospitalier Universitaire Charles Nicolle, Rouen; C. Adrie, M.D., M. Thuong, M.D., Medical–Surgical ICU, Hôpital Delafontaine, Saint Denis; J. Morel, M.D., G. Damian, M.D., Medical–Surgical ICU, Centre Hospitalier Universitaire, Saint Etienne; A. Bretsztajn, M.D., F. Bigorgne, R.N., Medical ICU, Hôpital Général, Salon de Provence; P. Sauder, M.D., Medical ICU, Hôpital Civil, Strasbourg; D. Hurel, M.D., H. Betefa, M.D., Medical Surgical ICU, Hôpital Foch, Suresnes; H. Margonari, M.D., Medical–Surgical ICU, Centre Hospitalier Thonon les Bains, Hôpitaux du Léman, Thonon les Bains; O. Angles, M.D., M. Genestal, M.D., Medical–Surgical ICU, Hôpital Purpan, Toulouse; J. Y. Petit, M.D., Medical–Surgical ICU, Hôpital Général, Valence; P. Rosamel, M.D., C. Combe, M.D., Medical–Surgical ICU, Hôpital Général, Villefranche; Z. Haddad, M.D., H. Salman, M.D., Medical Surgical ICU, Institut Gustave Roussy, Villejuif; P. Gaussorgues, M.D., Medical–Surgical ICU, Clinique du Tonkin, Villeurbanne, France.
OUTCOMEREA experts for the Delphi process: Phase 1: development of the list of medical errors: Dr. Christophe Adrie, Réanimation Polyvalente, Hôpital Delafontaine, Saint Denis; Dr. Thierry Boulain, Réanimation Polyvalente, Hôpital de la Source, Orléans; Dr. Marie Bruandet, Neurologie Vasculaire, Hôpital de la Pitié Salpétrière, Paris; Dr. Christian Brun Buisson, Réanimation Médicale, Hôpital Henri Mondor, Créteil; Dr. Alain Cariou, Réanimation Médicale, Groupe Hospitalier Cochin St Vincent de Paul, Paris; Dr. Bernard Clair, Réanimation Médicale, Hôpital Raymond Poincaré, Garches; Dr. De Cagny, Réanimation Nephrologique, CHU Amiens, Amiens; Dr. Bernard De Jonghe, Réanimation Médicale, Centre intercommunal, Poissy; Mr. Jean Paul Lamoureux, Réanimation Polyvalente, Hôpital Sainte Marguerite, Marseille; Dr. Philippe Loirat, Réanimation Médicale, Hôpital Foch, Suresnes; Mr. Louis Modestin, Réanimation Toxicologique, Hôpital Lariboisière, Paris; Dr. Laurent Papazian, Réanimation Médicale, Hôpital Sainte Marguerite, Marseille; Mr. Bruno Poupin, Réanimation Polyvalente, Centre Hospitalier les Oudaries, La Roche Sur Yon; Dr. Antoine Rabbat, Réanimation Médicale, Hôpital Hôtel Dieu, Paris; Dr. Bruno Raynard, Réanimation Polyvalente, Institut Gustave Roussy, Villejuif; Dr. Santhi Samy Modeliar, Réanimation Médicale, Centre Hospitalier Sud, Amiens; Dr. Fabienne Saulnier, Réanimation Médicale, Hôpital Calmette, Lille; Dr. Frédérique Shortgen, Réanimation Médicale, Hôpital Bichat, Paris; Dr. Antoine Vieillard Baron, Réanimation Médicale, Hôpital Ambroise Paré, Boulogne Billancourt, France.
Phase 2: Selection of medical errors using a Delphi process: Dr. Charles Arich, Réanimation Médicale, Centre Hospitalier Universitaire, Nîmes; Dr. Patrick Arnodo, Réanimation et Surveillance Continue, Centre Hospitalier Général, Aix en Provence; Dr. Jérome Baudot, Réanimation Polyvalente, Centre Hospitalier Général, Avignon; Mme. Simone Boehm, Unités des Grand Brûlés, CHU La Conception, Marseille; Mr. Frédéric Boissier, Réanimation Chirurgicale, Hôpital Bicêtre, Bicêtre; Dr. Eric Boulet, Réanimation Polyvalente, Centre Hospitalier, Pontoise; Mme. Ghylaine Bretteville, Réanimation Polyvalente, Hôpital Saint Joseph, Paris; Dr. Antoine Caubel, Pneumologie, Centre Hospitalier de Bretagne, Lorient; Dr. Julien Charpentier, Réanimation Polyvalente, Groupe Hospitalier Cochin St Vincent de Paul, Paris; Dr. Yaen Claessens, Urgences, Groupe Hospitalier Cochin St Vincent de Paul, Paris; Dr. Christophe Clec'h, Réanimation Médicale, Hôpital Avicenne, Avicenne; Mme. Laurence Colombet, Réanimation Médicale, Centre Hospitalier Albert Michallon, Grenoble; Dr. M. Darmon, Hôpital Nord, Saint Etienne, France; Dr. Nicolas Deye, Réanimation Médicale, Hôpital Lariboisière, Paris; Mme. Martine Dumain, Clinique Bizet, Paris; Dr. Dany Goldran Toledano, Hôpital Général, Gonesses; Dr. Alexandre Hertig, Réanimation Néphrologique, Hôpital Tenon, Paris; Dr. Jean Claude Lacherade, Réanimation Médicale, Centre Hospitalier Intercommunal, Poissy; Dr. Michel Mazerolles, Urgences, Hôpital Rangueil, Toulouse; Dr. Jacques Merrer, Réanimation Médicale, Centre Hospitalier Intercommunal, Poissy; Dr. Benoit Misset, Réanimation Polyvalente, Hôpital Saint Joseph, Paris; Dr. Ismael Mohammedi, Réanimation Médicale, Hôpital Edouard Herriot, Lyon; Dr. Laurent Montesino, Réanimation Polyvalente, Centre Hospitalier, Longjumeau; Dr. Delphine Moreau, Réanimation Médicale, Centre Hospitalier Montpied, Clermont Ferrand; Dr. Bruno Mourvillier, Réanimation Médicale et Infectieuse, Hôpital Bichat, Paris; Dr. Michel Pinsard, Réanimation Médicale, Hôpital Jean Bernard, Poitiers; Dr. Jean Philippe Rigaud, Réanimation Polyvalente, Centre Hospitalier, Dieppe; Dr. Carole Schwebel, Réanimation Médicale, Hôpital Albert Michallon, Grenoble; Dr. Jean Pierre Sollet, Réanimation Polyvalente, Centre Hospitalier Victor Dupouy, Argenteuil; Mr. Thierry Stella, Réanimation Polyvalente, Hôpital Ambroise Paré, Boulogne Billancourt; Dr. Fabienne Tamion, Réanimation Médicale, Centre Hospitalier Charles Nicolle, Rouen; Dr. Marie Thuong, Réanimation Médicale, Hôpital Delafontaine, Saint Denis; Mr. Daniel Tricoire, Réanimation Polyvalente, Centre Hospitalier Les Oudaries, La Roche Sur Yon; Dr. Gilles Troché, Réanimation Médico-Chirurgicale, Centre Hospitalier Mignot, Le Chesnay; Dr. Jean Ralph Zahar, Microbiologie et Unité D'Hygiène, Hôpital Necker, Paris, France.
| 1. | Leape LL, Brennan TA, Laird N, Lawthers AG, Localio AR, Barnes BA, Hebert L, Newhouse JP, Weiler PC, Hiatt H. The nature of adverse events in hospitalized patients: results of the Harvard Medical Practice Study II. N Engl J Med 1991;324:377–384. |
| 2. | Brennan TA, Leape LL, Laird NM, Hebert L, Localio AR, Lawthers AG, Newhouse JP, Weiler PC, Hiatt H. Incidence of adverse events and negligence in hospitalized patients: results of the Harvard Medical Practice Study I. N Engl J Med 1991;324:370–376. |
| 3. | Bates DW, Pruess K, Souney P, Platt R. Serious falls in hospitalized patients: correlates and resource utilization. Am J Med 1995;99:137–143. |
| 4. | Runciman WB. Lessons from the Australian Patient Safety Foundation: setting up a national patient safety surveillance system—is this the right model? Qual Saf Health Care 2002;11:246–251. |
| 5. | The Joint Commission. Sentinel Events [Internet]. Oakbrook Terrace, IL: The Joint Commission. Available at: http://www.jointcommission.org/SentinelEvents/ |
| 6. | Mayor S. English NHS to set up new reporting system for errors. BMJ 2000;320:1689. |
| 7. | Journal Officiel de la République Française arrêté du 25 Avril 2006 relatif aux modalités de l'expérimentation de déclaration des évènements indésirables graves liés des soins réalisés lors d'investigations, de traitements ou d'action de prévention autres que les infections nosocomiales. Available from: http://www.journal-officiel.gouv.fr/. |
| 8. | Garrouste-Orgeas M, Timsit JF, Adrie C, Tafflet M, Soufir L, Phillipart F, Zahar JR, Clec'h C, Toldgran-Toledano D, Jamali S, et al. Impact of adverse events in critically ill patients. Crit Care Med 2008;36:2041–2047. |
| 9. | Thomson R, Lewalle P, Sherman H, Hibbert P, Runciman W, Castro G. Towards an international classification for patient safety: a Delphi survey. Int J Qual Health Care 2009;2:9–17. |
| 10. | Chang A, Schyve PM, Croteau RJ, O'Leary DS, Loeb JM. The JCAHO patient safety event taxonomy: a standardized terminology and classification schema for near misses and adverse events. Int J Qual Health Care 2005;17:95–105. |
| 11. | Soufir L, Garrouste-Orgeas M, Vesin A, Tabah A, Dumay MF, Azoulay E, Carlet J, Timsit JF. Indicators of iatrogenic events in intensive care units: a French multicenter study [abstract]. Intensive Care Med 2007;33:S192. |
| 12. | Kohn LT, Corrigan JM, Donaldson MS, editors. To err is human: building a safer health system. Washington, D.C.: National Academies Press; 2000. |
| 13. | Knaus WA, Zimmerman JE, Wagner DP, Draper EA, Lawrence DE. APACHE—acute physiology and chronic health evaluation: a physiologically based classification system. Crit Care Med 1981;9:591–597. |
| 14. | Le Gall JR, Lemeshow S, Saulnier F. A new Simplified Acute Physiology Score (SAPS II) based on a European/North American multicenter study. JAMA 1993;270:2957–2963. |
| 15. | Osmon S, Harris CB, Dunagan WC, Prentice D, Fraser VJ, Kollef MH. Reporting of medical errors: an intensive care unit experience. Crit Care Med 2004;32:727–733. |
| 16. | Graf J, Von Den Driesch A, Koch KC, Janssens U. Identification and characterization of errors and incidents in a medical intensive care unit. Acta Anaesthesiol Scand 2005;49:930–939. |
| 17. | Giraud T, Dhainaut JF, Vaxelaire JF, Joseph T, Journois D, Bleichner G, Sollet JP, Chevret S, Monsallier JF. Iatrogenic complications in adult intensive care units: a prospective two-center study. Crit Care Med 1993;21:40–51. |
| 18. | De Lassence A, Alberti C, Azoulay E, Le Miere E, Cheval C, Vincent F, Cohen Y, Garrouste-Orgeas M, Adrie C, Troche G, et al. Impact of unplanned extubation and re intubation after weaning on nosocomial pneumonia risk in the intensive care unit: a prospective multicenter study. Anesthesiology 2002;97:148–156. |
| 19. | Lorente L, Huidobro MS, Martin MM, Jimenez A, Mora ML. Accidental catheter removal in critically ill patients: a prospective and observational study. Crit Care 2004;8:R229–R233. |
| 20. | De Lassence A, Timsit JF, Tafflet M, Azoulay E, Jamali S, Vincent F, Cohen Y, Garrouste-Orgeas M, Alberti C, Dreyfuss D. Pneumothorax in the intensive care unit: incidence, risk factors, and outcome. Anesthesiology 2006;104:5–13. |
| 21. | Lefrant JY, Muller L, De La Coussaye JE, Prudhomme M, Ripart J, Gouzes C, Peray P, Saissi G, Eledjam JJ. Risk factors of failure and immediate complication of subclavian vein catheterization in critically ill patients. Intensive Care Med 2002;28:1036–1041. |
| 22. | Boyle D, O'Connell D, Platt FW, Albert RK. Disclosing errors and adverse events in the intensive care unit. Crit Care Med 2006;34:1532–1537. |
| 23. | Paine LA, Baker DR, Rosenstein B, Pronovost PJ. The Johns Hopkins Hospital: identifying and addressing risks and safety issues. Jt Comm J Qual Saf 2004;30:543–550. |
| 24. | Nuckols TK, Bell DS, Liu H, Paddock SM, Hilborne LH. Rates and types of events reported to established incident reporting systems in two US hospitals. Qual Saf Health Care 2007;16:164–168. |
| 25. | Blendon RJ, DesRoches CM, Brodie M, Benson JM, Rosen AB, Schneider E, Altman DE, Zapert K, Herrmann MJ, Steffenson AE. Views of practicing physicians and the public on medical errors. N Engl J Med 2002;347:1933–1940. |
| 26. | Leape LL. Reporting of adverse events. N Engl J Med 2002;347:1633–1638. |
| 27. | Michel P, Quenon JL, De Sarasqueta AM, Scemama O. Comparison of three methods for estimating rates of adverse events and rates of preventable adverse events in acute care hospitals. BMJ 2004;328:199. |
| 28. | Rudis MI, Brandl KM; Society of Critical Care Medicine and American College of Clinical Pharmacy Task Force on Critical Care Pharmacy Services. Position paper on critical care pharmacy services. Crit Care Med 2000;28:3746–3750. |
| 29. | Brilli RJ, Spevetz A, Branson RD, Campbell GM, Cohen H, Dasta JF, Harvey MA, Kelley MA, Kelly KM, Rudis MI, et al. Critical care delivery in the intensive care unit: defining clinical roles and the best practice model. Crit Care Med 2001;29:2007–2019. |
| 30. | Leape LL, Cullen DJ, Clapp MD, Burdick E, Demonaco HJ, Erickson JI, Bates DW. Pharmacist participation on physician rounds and adverse drug events in the intensive care unit. JAMA 1999;282:267–270. |
| 31. | Vassal T, Anh NG, Gabillet JM, Guidet B, Staikowsky F, Offenstadt G. Prospective evaluation of self-extubations in a medical intensive care unit. Intensive Care Med 1993;19:340–342. |
| 32. | Rubins HB, Moskowitz MA. Complications of care in a medical intensive care unit. J Gen Intern Med 1990;5:104–109. |
| 33. | Huang DT, Clermont G, Sexton JB, Karlo CA, Miller RG, Weissfeld LA, Rowan KM, Angus DC. Perceptions of safety culture vary across the intensive care units of a single institution. Crit Care Med 2007;35:165–176. |
| 34. | Embriaco N, Azoulay E, Barrau K, Kentish N, Pochard F, Loundou A, Papazian L. High level of burnout in intensivists: prevalence and associated factors. Am J Respir Crit Care Med 2007;175:686–692. |
| 35. | Landrigan CP, Rothschild JM, Cronin JW, Kaushal R, Burdick E, Katz JT, Lilly CM, Stone PH, Lockley SW, Bates DW, et al. Effect of reducing interns' work hours on serious medical errors in intensive care units. N Engl J Med 2004;351:1838–1848. |
| 36. | Needham DM, Sinopoli DJ, Thompson DA, Holzmueller CG, Dorman T, Lubomski LH, Wu AW, Morlock LL, Makary MA, Pronovost PJ. A system factors analysis of “line, tube, and drain” incidents in the intensive care unit. Crit Care Med 2005;33:1701–1707. |
| 37. | Cullen DJ, Sweitzer BJ, Bates DW, Burdick E, Edmondson A, Leape LL. Preventable adverse drug events in hospitalized patients: a comparative study of intensive care and general care units. Crit Care Med 1997;25:1289–1297. |
| 38. | Reader TW, Flin R, Cuthbertson BH. Communication skills and error in the intensive care unit. Curr Opin Crit Care 2007;13:732–736. |
| 39. | Valentin A, Capuzzo M, Guidet B, Moreno RP, Dolanski L, Bauer P, Metnitz PG. Patient safety in intensive care: results from the multinational Sentinel Events Evaluation (SEE) Study. Intensive Care Med 2006;32:1591–1598. |
| 40. | Van Den Berghe G, Wouters P, Weekers F, Verwaest C, Bruyninckx F, Schetz M, Vlasselaers D, Ferdinande P, Lauwers P, Bouillon R. Intensive insulin therapy in the critically ill patients. N Engl J Med 2001;345:1359–1367. |
| 41. | Brunkhorst FM, Engel C, Bloos F, Meier-Hellmann A, Ragaller M, Weiler N, Moerer O, Gruendling M, Oppert M, Grond S, et al. Intensive insulin therapy and Pentastarch resuscitation in severe sepsis. N Engl J Med 2008;358:125–139. |
| 42. | Wiener RS, Wiener DC, Larson RJ. Benefits and risks of tight glucose control in critically ill adults: a meta-analysis. JAMA 2008;300:933–944. |
| 43. | Ferraris VA, Propp ME. Outcome in critical care patients: a multivariate study. Crit Care Med 1992;20:967–976. |
| 44. | Chacko J, Raju HR, Singh MK, Mishra RC. Critical incidents in a multidisciplinary intensive care unit. Anaesth Intensive Care 2007;35:382–386. |
| 45. | Ursprung R, Gray JE, Edwards WH, Horbar JD, Nickerson J, Plsek P, Shiono PH, Suresh GK, Goldmann DA. Real time patient safety audits: improving safety every day. Qual Saf Health Care 2005;14:284–289. |
| 46. | Ridley SA, Booth SA, Thompson CM. Prescription errors in UK critical care units. Anaesthesia 2004;59:1193–1200. |
