RESUMO
BACKGROUND: One third of patients admitted to an intensive care unit (ICU) will develop delirium. However, delirium is under-recognized by bedside clinicians without the use of delirium screening tools, such as the Intensive Care Delirium Screening Checklist (ICDSC) or the Confusion Assessment Method for the ICU (CAM-ICU). The CAM-ICU was updated in 2014 to improve its use by clinicians throughout the world. It has never been validated compared to the new reference standard, the Diagnostic and Statistical Manual of Mental Disorders 5th version (DSM-5). METHODS: We made a prospective psychometric study in a 16-bed medical-surgical ICU of a French academic hospital, to measure the diagnostic performance of the 2014 updated CAM-ICU compared to the DSM-5 as the reference standard. We included consecutive adult patients with a Richmond Agitation Sedation Scale (RASS) ≥ -3, without preexisting cognitive disorders, psychosis or cerebral injury. Delirium was independently assessed by neuropsychological experts using an operationalized approach to DSM-5, by investigators using the CAM-ICU and the ICDSC, by bedside clinicians and by ICU patients. The sensitivity, specificity, positive and negative predictive values were calculated considering neuropsychologist DSM-5 assessments as the reference standard (primary endpoint). CAM-ICU inter-observer agreement, as well as that between delirium diagnosis methods and the reference standard, was summarized using κ coefficients, which were subsequently compared using the Z-test. RESULTS: Delirium was diagnosed by experts in 38% of the 108 patients included for analysis. The CAM-ICU had a sensitivity of 83%, a specificity of 100%, a positive predictive value of 100% and a negative predictive value of 91%. Compared to the reference standard, the CAM-ICU had a significantly (p < 0.05) higher agreement (κ = 0.86 ± 0.05) than the physicians,' residents' and nurses' diagnoses (κ = 0.65 ± 0.09; 0.63 ± 0.09; 0.61 ± 0.09, respectively), as well as the patient's own impression of feeling delirious (κ = 0.02 ± 0.11). Differences between the ICDSC (κ = 0.69 ± 0.07) and CAM-ICU were not significant (p = 0.054). The CAM-ICU demonstrated a high reliability for inter-observer agreement (κ = 0.87 ± 0.06). CONCLUSIONS: The 2014 updated version of the CAM-ICU is valid according to DSM-5 criteria and reliable regarding inter-observer agreement in a research setting. Delirium remains under-recognized by bedside clinicians.
RESUMO
INTRODUCTION: Delirium is common in Intensive-Care-Unit (ICU) patients but under-recognized by bed-side clinicians when not using validated delirium-screening tools. The Confusion-Assessment-Method for the ICU (CAM-ICU) has demonstrated very good psychometric properties, and has been translated into many different languages though not into French. We undertook this opportunity to describe the translation process. MATERIAL AND METHODS: The translation was performed following recommended guidelines. The updated method published in 2014 including introduction letters, worksheet and flowsheet for bed-side use, the method itself, case-scenarios for training and Frequently-Asked-Questions (32 pages) was translated into French language by a neuropsychological researcher who was not familiar with the original method. Then, the whole method was back-translated by a native English-French bilingual speaker. The new English version was compared to the original one by the Vanderbilt University ICU-delirium-team. Discrepancies were discussed between the two teams before final approval of the French version. RESULTS: The entire process took one year. Among the 3692 words of the back-translated version of the method itself, 18 discrepancies occurred. Eight (44%) lead to changes in the final version. Details of the translation process are provided. CONCLUSIONS AND RELEVANCE: The French version of CAM-ICU is now available for French-speaking ICUs. The CAM-ICU is provided with its complete training-manual that was challenging to translate following recommended process. While many such translations have been done for other clinical tools, few have published the details of the process itself. We hope that the availability of such teaching material will now facilitate a large implementation of delirium-screening in French-speaking ICUs.