Publications

BACKGROUND

Patients with low-risk chest pain are frequently readmitted for evaluation of recurrent chest pain. It is unknown whether stress testing during the first admission for chest pain is a cost-effective means of reducing readmissions.

METHODS

Using a hospital administrative database, we conducted a retrospective cohort study of patients aged ≥18 years admitted under "observation status" to Baystate Medical Center between January 2007 and July 2009 for chest pain without acute coronary syndrome. We compared subsequent emergency department (ED) visits, readmissions, and costs within 1 year for patients who had a stress test at index admission to those who did not, adjusting for age, gender, race, insurance, and comorbidities.

RESULTS

The cohort included 3315 patients. Most (n = 2376, 71.7%) had a stress test during the index admission. Within 1 year, 256 (7.7%) patients returned to the ED at least once with chest pain. Of these, 112 (43.8%) were admitted during their first return visit. In the multivariable model, return visits for chest pain were negatively associated with previous stress testing (odds ratio [OR]: 0.6, 95% confidence interval [CI]: 0.5 to 0.9). Once in the ED, however, the risk of admission did not vary by stress test during index admission (OR: 0.8, 95% CI: 0.4 to 1.4). Overall costs, including index admission and follow-up visits for chest pain, were higher for patients with stress testing at index admission.

CONCLUSION

Inpatient stress testing reduced subsequent resource utilization in terms of ED visits and resultant readmissions, but the savings were not enough to offset the cost of initial testing.

We analyzed 265 urethral Neisseria gonorrhoeae specimens collected from symptomatic males at San Francisco's municipal sexually transmitted disease clinic, a participant in the Gonococcal Isolate Surveillance Project, during 2009. We used N. gonorrhoeae multiantigen sequence typing to describe characteristics of patients infected with common sequence type families. Specimens were classified into 6 homology-based families and 1 additional family of all other identified strains. Strain family results were combined with results of culture-based antibiotic sensitivity minimum inhibitory concentration, sociodemographic and behavioral risk data collected at the clinic, and presence or absence of the mosaic penicillin-binding protein 2 (penA) allele. Characteristics of patients were compared across strain families through the use of χ(2) statistics. Among men who have sex with men, strain distribution differed by those reporting receptive oral sex as their only urethral exposure (P = 0.04), by number of sex partners (P = 0.03), and by race/ethnicity (P < 0.001); there were no differences by age or human immunodeficiency virus status. Also, among men who have sex with men, strain family distributions differed for culture specimens with reduced susceptibility to a range of antibiotics, as well as with presence of the mosaic penA allele (all P < 0.001). The combination of molecular, phenotypic, and epidemiologic data on N. gonorrhoeae infection could help develop a more complete epidemiology of gonorrhea in the United States.