Search Annals: Advanced search

Podcast Transcript - July 17, 2007

[ Return to Podcast Home Page ]

Hello, and welcome to the Annals of Internal Medicine audio summary for our July 17, 2007 issue. I’m Michael Berkwits, Deputy Editor at Annals.

Our lead article this week is about changes in hospital mortality associated with implementation of residency work hours regulations.

Most listeners know that new work hour regulations for postgraduate trainees at US medical schools went into effect in July of 2003. The new rules limited workweeks to 80 hours or fewer averaged over 4 weeks, and they limited continuous duty to 24 hours, with up to 6 added hours allowed for the transfer of care and didactic activities. They also mandated a 10 hour rest period between working.

Anyone who did their internal medicine training prior to implementation of these rules has strong opinions about them, but studies of the effects of the changes have been limited in scope and outcome. Authors Kanaka Shetty and Jayanta Bhattacharya from the Veterans Affairs Palo Alto Health Care System and Stanford University used data from a national sample of US hospitals to determine if the change in work hours was associated with changes in mortality across a broad range of diagnoses. They used ICD9 and procedure codes to classify patients as having medical or surgical diagnoses, and they classified hospitals as teaching or nonteaching based on the presence of one or more residency programs. Then they compared mortality among medical and surgical patients at the teaching hospitals in the 30 months before and 18 months after implementation of the regulations.

The authors found a slight but statistically significant decrease in mortality after implementation of the new rules among patients with medical diagnoses. The effect appeared to be more pronounced among patients older than 80, and among patients with infectious disease diagnoses, such as sepsis and pneumonia. They did not detect any changes in mortality among surgical patients.

The authors conclude that the changes in work hours seem not to have been harmful, and they may even have been associated with a decline in mortality among medical patients. But they are appropriately circumspect in their conclusions. The “before-after” study design used in this article is the only way to study one-time historical changes, but it is a weak study design scientifically, because it can’t isolate the exposure of interest – in this case implementation of new work hour regulations –from the many other factors that may also have changed during the same time. Also, the authors’ classification of hospitals as teaching and non-teaching was crude, given that many patients in hospitals with residency programs are not cared for by medical or surgical residents. For these and other reasons, the authors call for further study of the long-term effect of the change in regulations on both graduate medical education and patient care.

Leora Horwitz and her coauthors from Yale University and the Mid America Heart Institute in Kansas City, MO report findings from a similar, smaller but more detailed study. They looked at changes in outcomes in the 12 months before and after implementation of the work hour regulations among patients cared for by medical residents, and attending physicians on a non-teaching hospitalist service, at Yale-New Haven Hospital.

They looked for but did not detect any change in mortality. They did find other favorable effects. ICU length-of-stay appeared to declined by a mean of about 2 days. The proportion of patients discharged to home or to rehabilitation (as opposed to a skilled nursing facility) increased by about 5%. And pharmacist interventions to prevent error were reduced by about 2 interventions per 100 patient days. They looked for but did not find any differences in overall length-of-stay, in 30 day readmission, or in adverse drug-drug interactions.

These authors, like those of the first article, also conclude that the changes in work hours did no harm, and may even have led to some improvements in outcomes.

However, in an editorial on the two articles, Lee Goldman and Nicholas Fiebach of Columbia University point out that the effects detected in the Yale-New Haven Hospital study were partly attributable to a deterioration in outcomes on the hospitalist service, which made the apparent differences favoring the teaching services greater than if care had remained the same or had improved among the hospitalists. And the editorialists make the larger and important point that even if the changes in work hours did no harm in the measured outcomes, we know little about how patients and their families experience the changes; how the changes affect nurses and other members of health care teams; and how they affect education and training. These effects would also be important to study and document, at least before any further regulations are instituted.

The second article in this week’s issue reports on trends in antibiotic use for Neisseria gonorrheae infection, and trends in antibiotic susceptibility in GC isolates among men seen at STI clinics in 25 to 30 US cities between 1988 and 2003.

In 1993, the CDC began recommending fluoroquinolones as first-line treatment for gonorrhea because of the rise of PCN resistance. The ability to treat GC with a single oral dose of a fluoroquinolone was welcomed by patients and providers, both of whom were happy to leave behind IM antibiotic injections. In this week’s article, lead author Susan Wang from the Centers for Disease Control confirmed that the proportion of men treated with penicillin fell from 39.5% to 0% between 1988 and 2003, and that the proportion of men treated with fluoroquinolones rose from 0% to 42%. However, perhaps just as predictably, the percentage of isolates resistant to ciprofloxacin and other fluoroquinolones rose from 0% in 1991 to 4.1% in 2003, while resistance to penicillin, peaked at 19.6% of isolates in 1991, and declined to 6.5% by 2003. Isolates resistant to fluoroquinolones were also frequently resistant to tetracycline and macrolides, an important observation because people treated for GC are often treated for Chlamydia with these other drugs. The good news was that ceftriaxone did not take any significant hit; very few isolates were resistant to it.

Based on these findings, the authors urge prudent use of fluoroquinolones for gonorrhea, and they urge discontinuation of the drugs wherever the prevalence of fluoroquinolone resistance is high. They clarify that good treatment alternatives include ceftriaxone 125 mg IM in a single-dose; cefixime 400 mg po; and spectinomycin 2 g IM, IF the patient does not have pharyngeal infection. Azithromycin 2 g in a single dose is also an option, but it is not recommended by the CDC because it is expensive and causes GI side effects.

This issue also contains US Preventive Services Task Force clinical guidelines on screening for chlamydia.

The US Preventive Services Task Force is convened and supported by the Agency for Healthcare Research and Quality, the health services research counterpart to the National Institutes of Health. It is charged by Congress with reviewing scientific evidence for clinical preventive services, and for developing evidence-based recommendations for the delivery of those services to the community. The Task Force issues, and Annals typically publishes, recommendations for preventive services targeting conditions of substantial burden in the US that are typically managed in primary care settings.

This week’s recommendations on Chlamydia screening are distilled into a beautiful Table on p. 131 of the print journal. To summarize the recommendations and the Table, the Task Force recommends that providers screen all sexually active women aged 24 years and younger, and those 25 and older if they are at increased risk for infection because of a history of chlamydial infection or STI, new or multiple sexual partners, inconsistent condom use, or a history of sex work. The strength of the recommendations vary for pregnant and non-pregnant women, as do the screening intervals. For women who are not pregnant, the Task Force acknowledges that the optimal interval for screening is not known, although the CDC recommends screening at least annually for women at increased risk. Pregnant women who are 24 or younger, and older women at increased risk should be screened at their first prenatal visit. Patients who continue to be at risk or who become newly at risk during the course of the pregnancy should be screened in their third trimester. Screening can be performed with nucleic acid amplication tests, performed on urine specimens or those obtained from vaginal swabs. For details of treatment, the Task Force refers readers to the CDC website at www.cdc.gov/STD/treatment.

Other papers supporting these Clinical Guidelines in the issue include the Task Forces’ systematic review of evidence about chlamydial screening that is new since its last 2001 recommendation; a description of the current processes of the Task Force in developing their evidence-based recommendations; and a guide to the new content and format of the Task Force’s recommendation.

Listeners more broadly interested in the US Preventive Services Task Force can also go to its website at www.preventiveservices.ahrq.gov Click the “Recommendations” link for a view of all of the Task Force’s current Clinical Guidelines.

Other articles in this week’s issue include an analysis of National Health and Nutrition Examination Survey or NHANES data, documenting a prevalence of chlamydial infection among people aged 14 to 39 in the US of 2.2%, with about equal prevalence in women and men; and also documenting that nearly half of people with N. gonorrheae infection also are infected with Chlamydia. And we have an Update in General Internal Medicine, the first of 10 Updates Annals will publish this year, based on the popular Update presentations given at the College’s annual Internal Medicine 2007 meeting held in April in San Diego.

For full details of these and all articles, please consult your print journal, or go to www.annals.org.

Subscribers to Annals can receive CME credits by answering quiz questions about trends in antibiotic susceptibility of N. gonorrheae, and the USPSTF chlamydia screening guidelines, at cme.annals.org. Just click the “My CME” link, to register or sign in.

That’s it for today. Send feedback, suggestions and comments about this audio summary and the journal to podcast{at}acponline.org.

Technical support for this summary was provided by Andrew Langman and Neil Kohl. Special thanks to Kevin Stahl, Dan Rosenthal, and all the folks at WHYY, public radio and television of Philadelphia, who helped produce this podcast.

Check back in 2 weeks for a complete summary of our regularly scheduled August 7, 2007 issue.

Have a great fortnight, and thanks for listening.

[ Return to Podcast Home Page ]