In This Issue of Archives of Internal Medicine FREE

A prospective multicenter study was conducted in 14 French intensive care units (ICUs), with all patients being screened for methicillin-resistant Staphylococcus aureus (MRSA) at ICU admission. Of the 2347 sampled patients, 162 (6.9%) were positive for MRSA, 54.3% of whom were detected through screening specimens only. Factors associated with MRSA carriage were age older than 60 years, prolonged hospital stay in transferred patients, history of hospitalization or surgery, and presence of open skin lesions in directly admitted patients. Only universal screening detected MRSA carriage with acceptable sensitivity. These results suggest that screening for MRSA on admission is useful to identify the imported cases and should be performed in all ICU-admitted patients.

Paltiel et al evaluated the impact of a computerized alert for hospitalized patients with serum potassium levels below 3 mEq/L. Using a previously validated computerized audit technique, they found that this intervention was associated with a 36.1% decrease in the proportion of patients in whom serum potassium level was not remeasured after an initial low value, a 28.6% decrease in failure to correct the serum potassium level to the normal range and a 17.2% decrease in discharge from hospital with subnormal potassium level. The authors conclude that a computerized alert is a simple and effective measure for improving the management of this serious electrolyte abnormality.

Chronic sleep deprivation is common in today's society. Nevertheless, the long-term health consequences of sleep deprivation are unclear. A cohort of 71 617 US female health professionals (aged 45-65 years) without reported coronary heart disease (CHD) at baseline were mailed a questionnaire in 1986 asking about daily sleep duration. Subjects were followed up until 1996 for the occurrence of CHD events. A total of 934 coronary events were documented (271 fatal and 663 nonfatal). Age-adjusted relative risks of CHD (with 8 hours of daily sleep being considered the referent group) for individuals reporting 5 or less hours of sleep, 6 hours of sleep, and 7 hours of sleep were 1.82 (95% confidence interval [CI], 1.34-2.41), 1.30 (95% CI, 1.08-1.57), and 1.06 (95% CI, 0.89-1.26), respectively. The relative risk for 9 or more hours of sleep was 1.57 (95% CI, 1.18-2.11). After adjusting for a variety of potential confounders, the relative risks of CHD for individuals reporting 5 or less hours of sleep, 6 hours of sleep, and 7 hours of sleep were 1.45 (95% CI, 1.10-1.92), 1.18 (95% CI, 0.98-1.42), and 1.09 (95% CI, 0.91-1.30), respectively. The relative risk for 9 or more hours of sleep was 1.38 (95% CI, 1.03-1.85). In this cohort of women, both short and long self-reported sleep durations were independently associated with a modestly increased risk of coronary events.

Although acetaminophen is recommended as first-line therapy for the pharmacological treatment of knee osteoarthritis, published data on its efficacy are scarce. Only one small study showed a statistically significant effect relative to placebo. This study compares acetaminophen, the nonsteroidal anti-inflammatory drug, diclofenac sodium, and placebo in 82 subjects with symptomatic medial knee osteoarthritis. Although diclofenac was effective in alleviating knee pain at 2 and 12 weeks, acetaminophen was indistinguishable from placebo at all time points (statistically significant differences are indicated in the Figure. Further, there was no apparent subset of patients who were more likely than others to derive symptomatic benefit from acetaminophen use. Given the data of this study and the limited data published earlier, the advocacy of acetaminophen use in the symptomatic treatment of knee osteoarthritis should be reconsidered.