Correspondance Return to March 6 , 2001 Table of Contents

J Miller highlights an important diagnostic challenge. If fever and an acute confusional state are the only presenting signs, when is urinary infection the cause of clinical deterioration in elderly residents of nursing homes? In this situation, although a positive urine culture is necessary to diagnose a urinary infection, it is not sufficient. At any given time as many as 50% of residents without symptoms have a positive urine culture, usually with pyuria, and a positive culture has a low predictive value for symptomatic urinary infection. Unfortunately, in the absence of localizing genitourinary findings such as costovertebral angle tenderness or hematuria, the relatively small proportion of these episodes that are due to urinary infection in the noncatheterized resident cannot be differentiated from episodes due to other causes. In the face of this uncertainty, the practitioner must base the treatment decision for each episode on his or her clinical judgement. The management issue here is not the treatment of asymptomatic bacteriuria, but the diagnosis of symptomatic urinary infection and the lack of specificity of that diagnosis. A major plea of my commentary is that physicians acknowledge this diagnostic uncertainty and consider a management approach of observation for residents who have only mild or moderate symptoms. In patients who are seriously ill, empiric antimicrobial therapy is certainly appropriate, given the diagnostic limitations. However, further systematic evaluation of diagnostic and management strategies in this population is necessary to identify optimal approaches to care.