antibiotic overprescribing

Two research letters appeared in JAMA specialty journals recently, reflecting antibiotic overprescribing.

1. the National Ambulatory Medical Care Survey and the National Hospital Ambulatory Medical Care Survey has lots of data on ambulatory care visits and includes patient demographics, ICD-9 diagnostic codes, and medications prescribed. this first report (see antibiotic overprescribing jama pedi 2014 in dropbox, or doi:10.1001/jamapediatrics.2014.1582) analyzes data from 1997-2011 for patients 3-17yo seen for sorethroat, and excluding visits with other infections present (eg cellulitis, which would require antibiotics), for a total of approx 12 million visits annually, finding:

--antibiotics prescribed for 60%, with 61% given narrow-spectrum antibiotics (penicillin, amoxacillin), and 39% given broad-spectrum (mostly macrolides, followed by second/third generation cephalosporins, then amox/clavulanate and first generation cephalosporins)

--over the 14 years of the study, scripts for narrow-spectrum antibiotics decreased from 65% to 52%, with increase in macrolides over time

--they conclude that likely inappropriate antibiotic prescribing, since only about 37% of pharyngitis is bacterial in children

so, this is a pretty powerful quick-and-dirty analysis suggesting overprescribing, with the attendant potential for adverse effects of antibiotics and development of resistance. but to me, perhaps the most shocking and disturbing aspect is the switch from narrow-spectrum to broad-spectrum antibiotics, for several reasons. first, macrolides (including azithro) are increasingly resistant to group A strep, with more than 10% of invasive isolates resistant (see a really great cdc report at http://www.cdc.gov/drugresistance/threat-report-2013 , which also documents a 3.4% resistance to clindamycin), so from a purely medical perspective, these drugs should only be given for patients who cannot tolerate penicillin. second, using these broad-spectrum antibiotics will create resistance for other organisms (eg, resistance of H Pylori to clarithro). and third, to continue to beat a dead horse, broad-spectrum antibiotics will cause more changes in the intestinal and other normal microbiota, with the potential for many untoward effects (see the slew of prior blogs on the microbiome). by the way, similar findings for  adults have found, with most getting broad-spectrum antibiotics (esp macrolides) as first-line therapy.

2. A study was done in the Partners system in Boston, using their electronic medical record, to assess the relationship between time-of-day and antibiotic prescriptions for acute respiratory infections, ARIs (see antibiotic time of day JAMA intmed 2014 in dropbox, or doi:10.1001/jamainternmed.2014.5225). they looked at adult patients aged 18-64 seen in one of 23 ambulatory sites for ARI and excluded those with chronic illnesses or other acute diagnoses. they also used national guidelines which define "antibiotics sometimes indicated" (otitis media, sinusitis, pneumonia, strep pharyngitis) and "antibiotics never indicated" (nonspecific URI, acute bronchitis, influenza, nonstreptococcal pharyngitis). results:

--21,867 visits to 204 clinicians, 44% resulting in antibiotic prescription

--there was a significant increase in antibiotic prescriptions in both the am and pm clinic sessions in the last 2 hours of the session vs the first 2 hours, with the 3rd hour having a significant 14% increase and the 4th hour a 26% increase as compared to the first hour

--65% of the visits were for ARI for which "antibiotics are never indicated": prescriptions for antibiotics ​increased from about 26% at 8am to about 33% at 11am, then decreased to about 31% at 1pm and increased to 34% at 4pm.

--35% of visits were for ARI for which "antibiotics are sometimes indicated: prescriptions for antibiotics increased from about 41% at 8am to 46% at 11am, then to 44% at 1pm increasing to 48% at 4pm.

so, this study suggests "decision fatigue", that the clinician near the end of the session is more likely to just give the antibiotics. not sure why. ?it takes too much time to talk with patients and explain why antibiotics are not necessary. ?clinician not want to get into yet another fight with a patient over antibiotics, so easier just to prescribe them. ?seeing too many patients and not want to devote the mental energy into the antibiotic discussion (eg, my guess is that there are more errors when a radiologist is reading their 80th chest xray in a session vs the first one, or a clinician who is plowing through a stack of EKGs probably makes more mistakes/pays less attention to the details of the 50th vs the 1st one of an afternoon). so, maybe we all need to take a 5-10 minute break every 1-2 hours, get a little exercise....  but again, the bottom line is that clinicians are still prescribing too many antibiotics (even fresh, early in the morning, clinicians are still prescribing antibiotics to 26% of patients where "antibiotics are never indicated").