NHAMCS ED Urinalysis, Blood Testing, and Length of Visit 2006-2015
Description
BACKGROUND: Despite the popular conception that ordering a urinalysis causes a significant increase in Emergency Department (ED) length of visit (LOV), there is little research on its actual impact . OBJECTIVES: This study investigated the quantitative impact of obtaining the results of a urinalysis, compared to the quantitative impact of obtaining the results of any laboratory testing of blood (“blood testing”), upon ED LOV in the National Hospital Ambulatory Medical Care Survey – Emergency Department [NHAMCS-ED] dataset. METHODS: The NHAMCS - ED dataset was queried from 2006 – 2015, comparing LOS in visits where urinalysis was ordered, blood testing was ordered, both were ordered, or neither. RESULTS: 1,232,279,000 ED visits with LOV data were found in the study period. Urinalysis was performed in 24.2% of visits, Blood testing in 36.7%, both in 18.2%, and none in 57.4%. Median LOV was 153 minutes. No blood or urine testing had a median LOV of 109, Urinalysis only 161 minutes, Blood testing only 221 minutes, and both 250 minutes. CONCLUSION: Urinalysis does increase LOV, but blood testing increases it more, with additive effects when both are ordered.
Files
Steps to reproduce
Primary datasets and documentation at https://www.cdc.gov/nchs/ahcd/datasets_documentation_related.htm ED Datasets from 2006 through 2015 were downloaded, and imported into a MySQL dataset via a custom script (https://www.phpclasses.org/package/10138-PHP-Extract-data-dictionaries-from-SAS-file-into-MySQL.html). The 2016 dataset, while available, was excluded as it did not have LOV data. Fields for Age, Gender, Immediacy (i.e. 5-level triage acuity), and LOV were collated from all 10 years into a single table. Of note, partway through the years, the age cap on reporting was increased from 93 to 100. All patients coded as “93 years and over” we recoded to 95 (the median age of patients between 93 and 100 in years where the age cap was 100). All patients coded at “100 years and over” we recoded to 100. All patients flagged with “Urinalysis” as “Yes” in the dataset were considered to have a urinalysis performed. For blood testing, given the shifting available options over the ten-year span, any patients that were flagged with “Complete Blood Count” or “Electrolytes” as “Yes” were considered to have blood testing performed. It was assumed by the investigators these panels would be included in almost all blood testing ordered.