Publications

BACKGROUND & AIMS

In the United States, upper gastrointestinal endoscopy is usually performed using intravenous sedation. Sedation increases the rate of both complications and costs of endoscopy. Unsedated esophagogastroduodenoscopy (EGD) using conventional 8-11-mm endoscopes is an alternative to sedated endoscopy but is generally perceived as unacceptable to many American patients. Unsedated EGD using ultrathin 5-6-mm endoscopes is better tolerated. A randomized trial comparing unsedated ultrathin EGD (UT-EGD) with sedated conventional EGD (C-EGD) in a diverse American population is needed.

METHODS

In this multicenter, randomized, controlled trial, 80 patients scheduled to undergo elective outpatient EGD were randomized to unsedated UT-EGD or sedated C-EGD. The study was carried out at San Francisco General Hospital, San Francisco Veterans Affairs Medical Center, and the Liver and Digestive Health Medical Clinic, San Jose.

RESULTS

Baseline characteristics of patients randomized to unsedated UT-EGD and sedated C-EGD were similar. Moreover, there were no significant differences in overall patient satisfaction and willingness to repeat endoscopy in the same manner among the 2 study groups. There was, however, a significant difference in median total procedure time between the 2 study groups of 1.5 hours (P < 0.0001). The mean (+/- SD) total procedure cost was 512.4 US dollars (+/- 100.8 US dollars) for sedated C-EGD and 328.6 US dollars (+/- 70.3 US dollars) for unsedated UT-EGD (P < 0.0001).

CONCLUSIONS

Patients undergoing unsedated UT-EGD are as satisfied as patients undergoing sedated C-EGD and are just as willing to repeat an unsedated UT-EGD. Unsedated UT-EGD was also faster, less costly, and may allow greater accessibility to this procedure.

The direct binding of Streptococcus mitis to human platelets is mediated in part by two proteins (PblA and PblB) encoded by a lysogenic bacteriophage (SM1). Since SM1 is the first prophage of S. mitis that has been identified and because of the possible role of these phage-encoded proteins in virulence, we sought to characterize SM1 in greater detail. Sequencing of the SM1 genome revealed that it consisted of 34,692 bp, with an overall G+C content of 39 mol%. Fifty-six genes encoding proteins of 40 or more amino acids were identified. The genes of SM1 appear to be arranged in a modular, life cycle-specific organization. BLAST analysis also revealed that the proteins of SM1 have homologies to proteins from a wide variety of lambdoid phages. Bioinformatic analyses, in addition to N-terminal sequencing of the proteins, led to the assignment of possible functions to a number of proteins, including the integrase, the terminase, and two major structural proteins. Examination of the phage structural components indicates that the phage head may assemble using stable multimers of the major capsid protein, in a process similar to that of phage r1t. These findings indicate that SM1 may be part of a discrete subfamily of the Siphoviridae that includes at least phages r1t of Lactococcus lactis and SF370.3 of Streptococcus pyogenes.