Chest Radiography Should Be Requested Only on Admission Based on Clinical Grounds.

OBJECTIVES: To determine the clinical utility and adverse consequences of routine admission chest x-ray (CXR) findings in patients with and without respiratory complaints and/or an abnormal chest examination. METHODS: In this prospective cohort study in an internal medicine department, we selected 273 patients and determined outcomes by chart review and physician interviews. The patients were divided into those with and without respiratory tract symptoms and/or findings on chest examination. The outcome variables were appropriate or inappropriate changes in treatment based on CXR findings. RESULTS: Of the 35 patients with respiratory tract symptoms/signs, 7 (20%) had a change in therapy based on CXR findings, which was effective in 5 of them. In the other 238 patients, an unexpected pleural empyema was detected in a hypotensive dialysis patient (0.4%, 95% confidence interval 0-2.3). Besides costs and radiation exposure, major adverse effects included two patients (0.8%, 95% confidence interval 0.1-3.0) with a false-positive test result that resulted in inappropriate hospitalizations and antibiotic therapy. In patients without respiratory tract symptoms or findings on physical examination, the clinical benefits and major adverse consequences were uncommon. CONCLUSIONS: Admission CXRs in patients without respiratory tract symptoms or findings are unwarranted except perhaps in older adult patients with comorbidities and an unclear admitting diagnosis.

The clinical utility and adverse consequences of the complete blood count in an internal medicine department.

The clinical utility and adverse consequences of the admission and follow-up complete blood count (CBC) in hospitalised patients are unclear. We selected 273 patients chosen from a single internal medicine department. To determine clinical utility and adverse consequences, we interviewed attending physicians and reviewed patients' charts. There were 12 (4.4%) patients hospitalised because of the CBC test result, six referred appropriately with a low haemoglobin concentration found in outpatient clinics and six (2.2%) patients (95% confidence interval 0.8-4.7%) inappropriately hospitalised because of incidental findings. In the hospital, according to the physicians, nearly all treatment changes made were for blood transfusions that were not indicated in 18 (6.6%) patients (95% confidence interval 4.0-10.2%). The only unexpected findings were in four patients with an indication for a blood transfusion admitted with an acute coronary syndrome and haemoglobin values 8-9.9 g/dL, and in one bedridden patient with dementia with acute myeloid leukaemia. There were 290 follow-up CBC tests not resulting in differential treatment. We conclude that admission CBC tests commonly lead to adverse consequences, due to physician errors in judgement. Incidental findings of anaemia justify CBC testing in patients with an acute coronary event. The rare patient with an incidental finding resulting in appropriate differential treatment might justify non-selective admission CBC counts, if physician education reduces the rate of inappropriate blood transfusions.