3: Prevalence and Prediction of Deterioration After Drowning In the Emergency Department

Article Outline

• Study Objective
• Methods
• Results
• Conclusion
• Copyright

Back to Article Outline

Study Objective 

Fears of decompensation in the well-appearing drowned patient present a unique challenge to the emergency physician. Lack of evidence-based guidelines result in prolonged emergency department (ED) observation and overnight hospital admissions. Our study sought to determine the prevalence of ED deterioration and clinical variables associated with deterioration for incorporation into a clinical prediction rule for early ED discharge.

Back to Article Outline

Methods 

This study is a retrospective cohort review of all pediatric drowning cases seen between July 2007 and June 2009 at a single urban tertiary care level 1 trauma pediatric ED. Cases were identified by computerized search utilizing the ICD-9 code (994.1) for drowning. We excluded cases not related to an acute drowning event, transferred from an outside ED, or admitted from our ED directly without observation. Two reviewers abstracted data on the following clinical variables from the hospital record: submersion time, out-of-hospital airway utilization, presenting ED mental status, presenting ED respiratory status, initial chest x-ray results, ED observation time, and ED disposition. The primary outcome measure was deterioration during ED observation requiring admission to the hospital. Association between predictor variable and outcomes was measured by univariate odds ratios and descriptive statistics. A clinical prediction rule was developed from these variables and the test characteristics of the rule were determined.

Back to Article Outline

Results 

176 cases were identified. 101 cases admitted directly without ED observation, 9 cases unrelated to an acute drowning event, and 21 transfers were excluded. This left a study population of 45 patients observed in the ED with 5 (11.1%) experiencing deterioration and subsequent admission. Mean ED observation time was 5 hours, (range: 0-14). Mean ED deterioration time was 2.9 hours, (range 0.5-4.25). No patients discharged had a return visit to an ED in our hospital system. Advanced EMS airway (OR 26, CI 3-195) was predictive of ED deterioration. Submersion time >1 min (OR 4, CI 0.63-35), abnormal pulmonary status (OR 2, CI 0.21-25), and abnormal chest x-ray (OR 4, CI 0.64-25) trended towards prediction of ED deterioration. Normal mental status was present in 4 (80%) cases of ED deterioration. A clinical decision rule comprised of these variables had a sensitivity of 80%, specificity 68%, PPV 24%, NPV 96%, LR(+) 3.08, and LR(-) 0.3.

Back to Article Outline

Conclusion 

This is the first study to determine prevalence of ED deterioration and associated predictors in the drowned patient. A prediction rule combining clinical variables known to the emergency physician showed discriminatory value in determining patients at risk for ED deterioration. Our study supports the need for a larger prospective study for further refinement and validation of this rule.