1). Exclusion Any concurrent comorbidities in the recorded diagnoses Follow-up visit of a prior ED visit Age<18 or
>64years The dependent variables were 1) the prescription of antibiotics and 2) the prescription of imaging studies (X-ray and CT). Several factors that may be associated with the prescription of antibiotics and imaging were investigated. Inhibitors,research,lifescience,medical The factors were selected a priori based on results from previous studies and the availability of information in the NHAMCS data. The types of URIs were categorized as URI NOS, nasopharyngitis, laryngitis, bronchitis, influenza, and multiple URI diagnoses. Nasopharyngitis, laryngitis and influenza were combined due to small cell sizes. Vital signs at presentation included whether the temperature was >100.4 Fahrenheit, whether the patient had tachydcardia (heart rate>100 beats per minute) Inhibitors,research,lifescience,medical and whether the patient had systolic blood Epigenetic inhibitor supplier pressure (SBP)>160mmHg or diastolic
blood pressure (DBP) >100mmHg. Less than 20 patients had bradycardia (hear rate<60 beats per minute) and pulse oximetry<92%. Due to the concern of small cell sizes, Inhibitors,research,lifescience,medical bradycardia and oxygen saturation were not examined. Characteristics of a visit included whether a patient presented with moderate or severe pain, waiting time to see a provider longer than 2hours, whether a physician saw the patient, and the season. Patient demographic characteristics examined were age (41–64 vs. 18–40years of age), sex, race, and ethnicity. Sources of payment were Inhibitors,research,lifescience,medical private insurance, Medicare, Medicaid, self-pay, and others. Sources of payment were not mutually exclusive because a patient may have multiple insurance types, for example, Inhibitors,research,lifescience,medical Medicare and private insurance. Geographic characteristics were Metropolitan Statistical Areas (MSA) and region (Northeast, Midwest, South and West). To achieve a sufficient sample size, 2007 and 2008
were combined, as done in previous studies of URIs using NHAMCS. The complex sampling design was controlled for in all analyses to provide nationally representative mafosfamide estimates. Statistical software SAS® (SAS Institute, Cary, NC) and Stata ® (StataCorp LP, College Station, TX) were used to perform the analyses. First, the prescribing patterns of antibiotics and imaging were estimated. The most frequently prescribed antibiotic classes were then identified. Two multivariate logistic regressions were performed for prescribing antibiotics and imaging studies, respectively, to identify the effect of each independent variable, controlling for the confounding factors.