Introduction
Defining safety
Measuring safety
-
The medical review. Reviews that do not target selected indicators are time-consuming and depend on the information available in the charts. Reviews can focus on selected indicators that can be assessed using the administrative data, discharge summaries, or mortality/morbidity review data. Medical reviews may be conducted manually or electronically using text words or text mining. Factors that may limit the use of the medical review method include absence of electronic medical records, paucity of resources for performing the reviews, variability in the terms used to label adverse events, and spelling mistakes. Failure to standardize the terminology may increase the difficulty of the search and the risk of false-positive results. Moreover, the analysis of documented adverse events requires considerable skill in interpreting the data. A meta-analysis comparing the rate of detection of pharmacists vs. nonpharmacists revealed a high level of adverse-event detection by pharmacists [16].
-
Voluntary reporting is the method most often used to detect medical errors and adverse events. Limitations include underreporting due to time constraints, lack of adequate reporting systems, fear of litigation, a reluctance to report one's own errors, uncertainty of the clinical importance of the events, and the lack of changes after reporting. However, this reporting method is the most useful for inducing behavioral changes, demonstrating the benefits of adverse-event reporting, and allowing us to learn from our errors. The presence of a multidisciplinary safety team might facilitate voluntary reporting.
-
Medical errors and adverse events also can be detected by direct observation at the bedside [17, 18]. This method is useful for detecting errors by omission. For example, medication errors can occur at any stage of the medication process (prescription, delivery, dispensing, administration, and monitoring), Medication error rates varied in the studies according to the definitions used, the medication process being evaluated, and the method of reporting. A pharmacist at the bedside can collect errors by omissions not detected by voluntary reporting. The medication error rate varied from 7.45/1,000 patient-days with voluntary reporting to 560/1,000 patient-days with daily routine observation of prescriptions [10, 12]. Similarly, the presence of a trained clinical research assistant who collected medical errors increased the rate from 2.2/1,000 to 597/1,000 patient-days in the IATROREF studies [14, 19].
-
The past several years have seen growing interest in learning from patients' experiences of care safety in all countries [20], with an older tradition in the United States and the United Kingdom via the CAHPS (Consumer Assessment of Healthcare Providers and Systems) and National Health Service (NHS), respectively. In 2007, the OECD (Organisation for Economic Co-operation and Development) established the patient's experience as a key priority. In the ICU, many patients are too ill to report on their own experience, but information can be obtained from families instead.
Process indicators |
---|
Mechanical ventilation
|
Overinflation of the endotracheal balloon [14] |
Sedation
|
Screening for ventilator weaning readiness [73] |
Sedation interruption [73] |
Sedation monitoring [73] |
Medication
|
Medication administered to wrong patient [14] |
Error administering anticoagulant medication [14] |
Error prescribing anticoagulant medication [14] |
Error administering vasoactive drugs [14] |
Error administering insulin [14] |
Death or serious disability associated with hypoglycaemia [75] |
IV lines
|
Screening for readiness for removal of central venous catheter [72] |
Management
|
Appropriate use of early enteral nutrition [72] |
Early management of severe sepsis, septic shock [72] |
Surgical intervention in traumatic brain injury with subdural and/or epidural brain trauma [72] |
Monitoring of intracranial pressure in severe traumatic brain injury with abnormal CT findings [72] |
Delay in surgical treatment [14] |
Change of route for quinolones IV/PO [72] |
Screening for MRSA on admission [76] |
Pain management in un sedated patients [72] |
Events during ICU transport [73] |
Complications
|
Pneumonia associated with mechanical ventilation [72] |
Accidental removal of a central venous catheter |
Catheter-related bloodstream infections [76] |
Death or serious disability associated with intravascular air embolism [75] |
Fall [14] |
Death or serious disability associated with a haemolytic reaction due to the administration of ABO-incompatible blood or blood product [75] |
Percentage of resistant organisms [74] |
Pressure sores [73] |
Outcome indicators
|
ICU mortality rate [74] |
Hospital mortality rate [73] |
Percentage of ICU patients with ICU stays longer than 7 days [74] |
Mean ICU length of stay [74] |
Mean days on mechanical ventilation [74] |
Rate of re-admissions < 72 hours [73] |
Family satisfaction [73] |
Structural indicators
|
Institutional variables
|
Process for ensuring staff competencies |
Transitional period to integrate new healthcare workers |
Clear task identification |
Absenteeism, magnitude of personnel turn-over |
Adverse-event reporting system |
Task variables
|
Availability of protocols |
Policy to prevent medication errors |
Policy to register outcomes |
Team variables
|
Adequacy of staffing |
Nurse-to-patient ratio |
Availability of an intensive care practitioner 24 h a day |
Pharmacist present during ICU rounds [77]) |
Communication or conflicts among team members [78] |