Nicola Mumoli1, Josè Vitale2, Alberto Pagnamenta2, Daniela Mastroiacovo3, Marco Cei4, Fulvio Pomero5, Matteo Giorgi-Pierfranceschi6, Lucia Giuntini7, Cesare Porta8, Riccardo Capra8, Antonino Mazzone8, Francesco Dentali9. 1. Department of Internal Medicine, Livorno Hospital, Livorno, Italy; Department of Internal Medicine, ASST Ovest Milanese, Magenta, Italy. Electronic address: nimumoli@tiscali.it. 2. Department of Intensive Medicine, Intensive Care Unit and Biostatistics Unit, Regional Hospital Mendrisio, Ente Ospedaliero Cantonale, Switzerland. 3. Angiology Unit, "SS Filippo and Nicola" Hospital, Avezzano, Italy. 4. Department of Internal Medicine, Cecina Hospital, Cecina, Italy. 5. Department of Internal Medicine, Michele e Pietro Ferrero Hospital, Verduno, Italy. 6. Department of Internal Medicine, Cremona Hospital, Cremona, Italy. 7. Department of Internal Medicine, Livorno Hospital, Livorno, Italy. 8. Department of Internal Medicine, ASST Ovest Milanese, Magenta, Italy. 9. Department of Clinical and Experimental Medicine, Insubria University, Varese, Italy.
Abstract
BACKGROUND: Chest radiography is universally accepted as the method of choice to confirm correct positioning of a nasogastric tube (NGT). Considering also that radiation exposure could increase with multiple insertions in a single patient, bedside abdominal ultrasound (BAU) may be a potentially useful alternative to chest radiography in the management of NGTs. RESEARCH QUESTION: What is the accuracy of BAU in confirming the correct positioning of an NGT? STUDY DESIGN AND METHODS: After a specific course consisting of 10 h of training, the authors studied, in a prospective multicenter cohort, the validity of BAU to confirm correct NGT placement. All patients were also evaluated by auscultation (whoosh test) and by chest radiography. Every involved operator was blind to each other. Interobserver agreement and accuracy analyses were calculated. RESULTS: This study evaluated 606 consecutive inpatients with an indication for NGT insertion. Eighty patients were excluded for protocol violation or incomplete examinations and 526 were analyzed. BAU was positive, negative, and inconclusive in 415 (78.9%), 71 (13.5%), and 40 (7.6%), respectively. The agreement between BAU and chest radiography was excellent. Excluding inconclusive results, BAU had a sensitivity of 99.8% (99.3%-100%), a specificity of 91.0% (88.5%-93.6%), a positive predictive value of 98.3% (97.2%-99.5%), and a negative predictive value of 98.6% (97.6%-99.7%). The accuracy of BAU slightly changed according to the different assignments of the uncertain cases and was improved by the exclusion of patients with an altered level of consciousness. INTERPRETATION: These results suggest that BAU has a good positive predictive value and may confirm the correct placement of NGTs when compared with chest radiography. However, considering its suboptimal specificity, caution is necessary before implementing this technique in clinical practice.
BACKGROUND: Chest radiography is universally accepted as the method of choice to confirm correct positioning of a nasogastric tube (NGT). Considering also that radiation exposure could increase with multiple insertions in a single patient, bedside abdominal ultrasound (BAU) may be a potentially useful alternative to chest radiography in the management of NGTs. RESEARCH QUESTION: What is the accuracy of BAU in confirming the correct positioning of an NGT? STUDY DESIGN AND METHODS: After a specific course consisting of 10 h of training, the authors studied, in a prospective multicenter cohort, the validity of BAU to confirm correct NGT placement. All patients were also evaluated by auscultation (whoosh test) and by chest radiography. Every involved operator was blind to each other. Interobserver agreement and accuracy analyses were calculated. RESULTS: This study evaluated 606 consecutive inpatients with an indication for NGT insertion. Eighty patients were excluded for protocol violation or incomplete examinations and 526 were analyzed. BAU was positive, negative, and inconclusive in 415 (78.9%), 71 (13.5%), and 40 (7.6%), respectively. The agreement between BAU and chest radiography was excellent. Excluding inconclusive results, BAU had a sensitivity of 99.8% (99.3%-100%), a specificity of 91.0% (88.5%-93.6%), a positive predictive value of 98.3% (97.2%-99.5%), and a negative predictive value of 98.6% (97.6%-99.7%). The accuracy of BAU slightly changed according to the different assignments of the uncertain cases and was improved by the exclusion of patients with an altered level of consciousness. INTERPRETATION: These results suggest that BAU has a good positive predictive value and may confirm the correct placement of NGTs when compared with chest radiography. However, considering its suboptimal specificity, caution is necessary before implementing this technique in clinical practice.