Predicting Treatment Failure in AOM

Like most infectious diseases, acute otitis media generally breaks down into three cohorts. There are viral infections, for which early antimicrobial therapy is virtually, by definition, unhelpful. Then, there are true bacterial infections – many of which resolve without substantial morbidity regardless of antimicrobial treatment, and those which require antimicrobial therapy to prevent such. The trick, and where modern medicine typically fails miserably, is rapidly predicting into which of these cohorts a patient may fall – a conundrum leading to the epidemic of antibiotic overuse.

This is a secondary analysis of a pediatric AOM trial, first published in the New England Journal of Medicine, looking at which patients were more likely to potentially fail conservative treatment. The intervention arm received amoxicillin/clavulanate, and treatment failure occured in 31.7% of children – vastly favoring the antibiotic arm – 44.9% vs. 18.6%. In theory, this exaggerated treatment effect might help better illuminate any small predictors – but, unfortunately, with only 319 patients, meaningful statistical significance on this data dredge is hard to come by. Worse still, the best predictor of treatment failure (or, really, lack thereof)? A peaked tympanogram (A and C curves) – you know, because we’re all routinely measuring tympanometry. Grossly bulging tympanic membranes were predictive of treatment failure, which has some face validity, at least – but, again, this is as compared between severe, moderate, and mild, which requires pneumatic otoscopy to differentiate.

The question here primarily concerns: can you take away good conclusions from bad data? The magnitude of the treatment effect seen in this trial far exceeded the treatment effect expected from antibiotics in other trials. And, consistent with that questionable generalizability, their findings reflect the stringent criteria determining their diagnosis of AOM. Then, they are relying upon their misguided definition for treatment failure, which relies on otoscopic signs, the same ones that will be colinear with worsened disease on initial examination. Unfortunately, the net result of all of this meandering is essentially no clinically useful insight. Considering the limitations the examination of the screaming ill toddler, more pragmatic approaches are necessary.

“Prognostic Factors for Treatment Failure in Acute Otitis Media”

Now It’s Fluids that Matter in Sepsis?

A few weeks ago, there was an article in the New England Journal of Medicine that dredged a retrospective data set to generate an association between timeliness different elements of a sepsis bundle and outcomes. In their analysis, antibiotics, but not fluid administration, was associated with a mortality increase. This has, at least, face validity – although, the association between timely blood cultures and serum lactate a little less so.

Now, conversely, we have another sepsis registry review attempting to tie time to fluid administration to mortality. This quality improvement registry prospectively identified patients with sepsis – and retrospectively abstracted their clinical data – between 2014 and 2016, resulting in a database of 11,182. In their analysis, mortality for patients receiving their first crystalloid within 30 minutes or within 30-120 minutes was ~18%, while mortality for patients whose fluids were initiated beyond the 120 minute limit was 24.5%.

Again, however, because these are comparisons performed on observational data, it is still subject to the slings and arrows of unmeasured confounders. Most patients whose fluid administration was started early had their care initiated in the Emergency Department – and, in clearly co-linear processes, had major elements of their care completed appropriately. This included repeat lactate measurements, antibiotics within 180 minutes of time zero, and, not only IVF within 120 minutes, but frankly, any IVF at at all. Nearly 60% of patients analyzed for their >120 minute cohort received <5 mL/kg or zero IVF in their first six hours from measurement time zero.

This is, probably, another study just cherry picking out one single feature of an entire process predicated on timely identification and treatment of sepsis. These patients did not simply have a mortality advantage because of the timeliness of IVF – it ties in to all aspects of care and attention given sepsis patients properly identified. The effect size here is probably less associated with delays just in IVF, but a comprehensive delay in diagnosis – and all its associated therapeutic misadventures.

“Patterns and Outcomes Associated With Timeliness of Initial Crystalloid Resuscitation in a Prospective Sepsis and Septic Shock Cohort”

Conjunctivitis: No Antibiotics, Please!

It’s the sad state of modern medicine – choose a common ambulatory condition, and you can find widespread avoidable overuse and waste. There is a spectrum of acceptability to this practice variation, of course, depending on the severity of consequences for missed or delayed diagnoses – but, for the most part, we’re just setting our professional respectability aflame.

This is a simple retrospective review of prescriptions associated with diagnoses of acute conjunctivitis. These authors reviewed records from a large managed care network and identified 340,372 patients with a clinical visit coded for acute conjunctivitis. Within 14 days of this visit, 58% of patients filled prescriptions for topical ophthalmologic medications. Considering most conjunctivitis encountered in the clinical setting is viral or allergic, obviously, the vast majority of these are wholly unnecessary. Then, frankly, while topical antibiotics mildly hasten the improvement of bacterial conjunctivitis, it is still a generally self-limited condition.

Ophthalmologists and optometrist visits were the least likely to have an antibiotic prescription associated with a visit for acute conjunctivitis, but 36% and 44%, respectively. Urgent Care Physicians and “Other Provider” – probably inclusive of Emergency Medicine – were at 68% and 64%, respectively. Fluoroquinolones accounted for 33% of antibiotic prescriptions – which is fabulous, because they are typically the most costly, and result in both increased risk for antimicrobial resistance and S. aureus endophthalmitis. Then, one in five prescriptions were for combination corticosteroid-antibiotic combination products – which are contraindicated, as they can prolong viral infections or worsen an underlying herpes simplex infection.

The American Academy of Ophthalmology contribution to Choosing Wisely recommends avoiding antibiotic prescriptions for viral conjunctivitis, and deferring immediate antibiotic therapy when the cause of conjunctivitis is unknown. Stop the madness! Everyone!

“Antibiotic Prescription Fills for Acute Conjunctivitis among Enrollees in a Large United States Managed Care Network”

I&D Alone or With Antibiotics for the Little Guys

Most physicians provide adjunctive antibiotic therapy for large abscesses following incision and drainage – the sorts where you need a bucket and a hose. Less clear has been the small abscess – but, in the age of MRSA, the fear factor has led many to cover these, regardless. Recent evidence suggests there is a small absolute benefit to antibiotic use and clinical cure, with an NNT around 14, along with other apparent benefits regarding re-infection and spread to household contacts. These trials, however, still enrolled patients with abscesses much larger than typically encountered in routine practice.

This trial is specifically designed to break the glass on “smaller skin abscesses” – just like in the title! What does small mean to these authors? It means a suppurative cavity of 5cm in diameter, or, up to the size of a cupcake:
abscess cakeSo, before we even start, we can see we may end up with issues regarding generalizability to many of the abscesses we encounter in the Emergency Department.

This trial is comprised of three arms – clindamycin, trimethoprim-sulfamethoxazole, and placebo – and enrolled 786 patients in an attempt to detect a 10% difference between arms while accounting for 20% attrition rate. The primary outcome was test of cure at 10 days after therapy, with a variety of secondary outcomes, including new infections at one month and treatment-related adverse events.

The winner, if one can be crowned, was not placebo. At the test of cure visit in the intention-to-treat population – and likewise, the population that could be evaluated – placebo lagged behind both clindamycin and TMP-SMX by approximately a 12% absolute magnitude of difference. Recurrent infections at the same site, or another site, were lowest in the clindamycin group at 6.8% – and similar between TMP-SMX and placebo, at 13.5% and 12.4%, respectively. However, clindamycin was implicated in the highest rate of adverse events, at 21.9%, compared with TMP-SMX and placebo, at 11.1% and 12.5%, respectively. Most of the difference in adverse events can be attributed to diarrhea illness, although clostridium difficile was not isolated in any cases. There was one case of systemic hypersensitivity reaction thought to be related to TMP-SMX.

There were two main drivers for the difference in test of cure between the placebo cohort and the two antibiotic cohorts, and these were use of rescue antibiotics during the follow-up period and new infections at another site. The use of rescue antibiotics is not necessarily a reliable measure of treatment failure, but it is still reasonable to suggest this difference would not arise by chance alone, despite the small sample. Regarding generalizability to practice, the minority of abscesses were cupcake-sized, but these were still fairly substantial infections. The median size of the abscess was about 2.2cm in diameter, with surrounding erythema of 5.9cm in greatest dimension.

The takeaway, then, hinges on the generalizability of their population to your individual patient. If these are “smaller” skin abscesses, then I wager the bulk of my abscess encounters are for “tinier” abscesses. I doubt this changes much current practice with regard to antibiotics, or antibiotic selection, for those treating abscesses in the 2+cm range, but I expect the differences in cure rates shrink for smaller lesions. It falls within the realm of acceptable practice variation to weigh the harms of antibiotic use with the chance of recurrence or new infection for those lesions.

“A Placebo-Controlled Trial of Antibiotics for Smaller Skin Abscesses”

Double Coverage, Cellulitis Edition

The Infectious Disease Society Guidelines are fairly reasonable when it comes to cellulitis. Non-suppurative cellulitis – that is to say, without associated abscess or purulent drainage – is much less likely to be methicillin-resistant s. aureus. The guidelines, therefore, recommend monotherapy with a ß-lactam, typically cephalexin. Conversely, with a suppurative focus, trimethoprim-sulfamethoxazole monotherapy is an appropriate option. However, it’s reasonable to estimate current practice involves prescribing both agents somewhere between one fifth and a quarter of cases – presumably both wasteful and potentially harmful. This trial, therefore, examines this practice by randomizing patients to either double coverage or cephalexin plus placebo.

The short answer: no difference. The rate of clinical cure was a little over 80% of both cohorts in the per-protocol population. Of those with follow-up and treatment failure, over half progressed to abscess or purulent drainage on re-evaluation – and about two-thirds were cultured out as s. aureus. There was no reliable evidence, however, co-administration of TMP-SMX prevented this progression.

The really fun part of this article, however ties into the second line of their abstract conclusion:

“However, because imprecision around the findings in the modified intention-to-treat analysis included a clinically important difference favoring cephalexin plus trimethoprim-sulfamethoxazole, further research may be needed.”

This hedging stems from the fact 17.8% were excluded from the enrolled cohort for inclusion in the per-protocol analysis – and, depending on the modified intention-to-treat analysis definition, there was actually up to a 7.3% difference in failure rate favoring double coverage (76.2% vs 69.0%). This resulted from almost twice as many patients in the cephalexin monotherapy cohort taking <75% of antimicrobial therapy, missing follow-up visits, or other protocol deviations.

The best Bayesian interpretation of this finding is probably – and this is where frequentism falls apart – simply to ignore it. The pre-study odds of dramatic superiority of double coverage are low enough, and the outcome definition for the modified intention to treat cohort in question is broad enough, this finding should not influence the knowledge translation of this evidence. Stick with the IDSA soft-tissue guidelines – and one antibiotic at a time, please.  It is important to recognize – and educate patients – that about 1 in 6 may fail initial therapy, and these failures to not necessarily reflect inappropriately narrow antibiotic coverage nor therapeutic mismanagement.

“Effect of Cephalexin Plus Trimethoprim-Sulfamethoxazole vs Cephalexin Alone on Clinical Cure of Uncomplicated Cellulitis”

Making Urine Cultures Great Again

As this blog covered earlier this month, the diagnosis of urinary tract infection – as common and pervasive as it might be – is still fraught with diagnostic uncertainty and inconclusive likelihood ratios. In practice, clinicians combine pretest likelihood, subjective symptoms, and the urinalysis to make a decision regarding treatment – and invariably err on the side of over-treatment.

This is an interesting study taking place in the Nationwide Children’s Hospital network regarding their use of urine cultures. In retrospect, these authors noted only half of patients initially diagnosed with UTI had the diagnosis ultimately confirmed by contemporaneous urine culture. Their intervention, then, in order to reduce harm from adverse effects of antibiotics, was to contact patients following a negative urine culture result and request antibiotics be stopped.

This tied into an entire quality-improvement procedure simply to use the electronic health record to accurately follow-up the urine cultures, but over the course of the intervention, 910 patients met inclusion criteria. These patients were prescribed a total of 8,648 days of antibiotics, and the intervention obviated 3,429 (40%) of those days. Owing to increasing uptake of the study intervention by clinicians, the rate of antibiotic obviation had reached 61% by the end of the study period.

There are some obvious flaws in this sort of retrospective reporting on a QI intervention, as there was no reliable follow-up of patients included. The authors report no patients were subsequently diagnosed with a UTI within 14 days of being contacted, but this is based on only 46 patients who subsequently sought care within their healthcare system within 14 days, and not any comprehensive follow-up contact. There is no verification or antibiotics actually being discontinued following contact. Then, finally, antibiotic-free days are only a surrogate for a reduction the suspected adverse events associated with their administration.

All that said, this probably represents reasonable practice. Considering the immense frequency with which urine cultures are sent and antibiotics prescribed for dysuria, the magnitude of effect witnessed here suggests a potentially huge decrease in exposure to unnecessary antibiotics.

“Urine Culture Follow-up and Antimicrobial Stewardship in a Pediatric Urgent Care Network”

Another Step in Antibiotics for Appendicitis

Antibiotics are unnecessary! No, antibiotics are great! No, we give too many antibiotics! It’s getting hard to keep track of which conditions we’re giving and withholding antibiotics for these days.

This article is a teaser for more evidence to come regarding strategies for managing appendicitis without surgical intervention. We’ve seen a few trials already, with essentially unconvincing results in either direction. A large trial regarding an antibiotics-first strategy in an adult population was criticized for using open surgical technique rather than laproscopic – and the one-year failure rate was still rather high. However, a pilot report in a pediatric population probably demonstrates an antibiotic-first strategy is still a reasonable option to present in shared decision-making.

This is a pilot project describing the initial results and feasibility outlook for an antibiotics-first protocol for appendicitis. In this protocol, patients randomized to an antibiotics-first strategy received an intravenous dose of ertapenem in the Emergency Department, were eligible for discharge directly from the Emergency Department, returned for a second dose of ertapenem the next day, and then completed an 8-day course of oral cefdinir and metronidazole.

In their pilot, 42 patients were screened and 30 patients consented for randomization. Of these, 15 were adults and 1 was a pediatric patient. Of the 15 adults, 14 felt well enough for discharge after initial Emergency Department observation. The pediatric protocol called for in-hospital observation regardless of symptoms at presentation.

The results are generally of lesser consequence than the effectiveness of this pilot demonstrating the feasibility of the protocol, and the yield at which patients could be enrolled for a larger trial. There were a couple instances of recurrent appendicitis in the antibiotics-first cohort, one of which was successfully treated with antibiotics a second time. There were a couple surgical complications in the surgery cohort. Costs and overall quality of life scores favored the antibiotics-only group, obviously – but, again, this sample is small enough none of these outcomes have been measured with reliable accuracy or precision.

I think it is reasonable to expect an antibiotics-first strategy to eventually take root as part of acceptable medical practice. However, I suspect this transition will be slow in coming – and more data would be quite helpful in determining any specific risks for antibiotic strategy failures.

“Antibiotics-First Versus Surgery for Appendicitis: A US Pilot Randomized Controlled Trial Allowing Outpatient Antibiotic Management”

Insight Is Insufficient

In this depressing trial, we witness a disheartening truth – physicians won’t necessarily do better, even if they know they’re not doing well.

This study tested a mixed educational and peer comparison intervention on primary care physicians in Switzerland, with an end goal of improving antibiotic stewardship for common ambulatory complaints. The “worst-performing” 2,900 physicians with respect to antibiotic prescribing rates were enrolled and randomized to the study intervention or none. The study intervention consisted of materials regarding appropriate prescribing, along with personalized feedback regarding where their prescribing rate ranked compared to the entire national cohort. The core of their hypothesis involved whether just this passive knowledge regarding their peer performance would exert normalizing influence over their practice.

Unfortunately, despite providing these physicians with this insight, as well as tools for improvement, the net effect of their intervention was effectively zero. There were some observations regarding changes in prescribing rates for certain age groups, and for certain types of antibiotics, but dredging through these secondary outcomes leads to only unreliable conclusions.

This is not particularly surprising data. These sorts of passive feedback mechanisms unhitched from material consequences have never previously been shown to be effective. There are other, more effective mechanisms – focused education, decision-support interventions, and shared decision-making – but, for a fragmented, national health system, this represented a relatively inexpensive model to test.

Try again!

“Personalized Prescription Feedback Using Routinely Collected Data to Reduce Antibiotic Use in Primary Care”

No, All Bacteria Do Not Require Antibiotics

The natural world is replete with bacteria.

Humans have existed on this planet for millennia.

In the ages before antibiotics, many humans succumbed to bacterial infections – while, of course, the vast majority survived.

This is not a profoundly reliable observational study, but it does help reinforce this basic concept. This report is a secondary analysis of the GRACE-10 study, which involved primary care patients recruited with a diagnosis of acute cough. The original study was a randomized, placebo-controlled trial for non-specific lower respiratory tract infection, as part of a genomics analysis for evaluation of antibiotic resistance.

This analysis, however, looks solely at the placebo arm, and examines the symptom course and resolution of those who were ultimately diagnosed with a bacterial cause of their LRTI and compares the with those who were not. Of the 834 patients included in their analysis (those with complete symptom diaries), 162 were thought to have a bacteria pathogen based on respiratory culture, nasal swab, or whole blood antibody titers.

S pneumoniae and H influenzae were the most common bacterial pathogens, with most of the remainder the “atypicals” for community-acquired pneumonia. And, at the end of the day: virtually everyone did fine. Patients with a confirmed bacterial pathogen in the setting of their LRTI improved slightly more slowly than those without, had more re-visits in follow-up due to worsening or new symptoms, and a greater percentage were placed on antibiotics in follow-up (12% vs. 6%). The remainder eradicated their bacterial pathogens without antibiotics – you know, the way humans and other contemporary mammals survived for eons.

Now, some of these cases positive for LRTI may be colonization and not pathogenic infection, while some of the negative cases were not diagnosed due to lack of sensitivity. But, regardless, the overall point of this article is probably valid – some bacterial infections will worsen, but in the generally healthy population, a delayed-antibiotic strategy might be valid as an attempt to improve antibiotic stewardship.

“Disease Course of Lower Respiratory Tract Infection With a Bacterial Cause”

More Coverage of Inappropriate Antibiotic Prescribing

If this feels like déjà vu, it might be because it is.

This short research letter in JAMA Internal Medicine describes patterns of antibiotic prescribing for three common conditions: otitis media, sinusitis, and pharyngitis. In all of these cases – in the infrequent occasion antibiotics are necessary – the appropriate first-line antibiotic is amoxicillin/penicillin. These authors estimate, based on treatment failures, allergies, and complicated disease, approximately 80% of antibiotic prescriptions for these conditions should be the first-line agents.

How did we do? Well, better in pediatrics than adults, but first-line prescribing ranged from a low of 37% to a high of 67%. The most commonly used inappropriate antibiotics were macrolides (invariably azithromycin) and fluoroquinolones. Macrolides are usually inappropriate due to high levels of resistance among common pathogens, and fluroquinolones are simply too broad-spectrum to be appropriate.

The catch, unfortunately, is the data source: the National Ambulatory Medical Care Survey, warts and all, from 2010 to 2011. The authors state they expect practice patterns have not changed much in the last five years, but it’s still a little challenging to generalize this to current practice.

Finally, as a nice corollary, this Medical Letter article was featured in JAMA regarding fluoroquinolones and their increasingly detected serious adverse effects. When antibiotics are truly necessary, physicians should try and choose one of the many alternatives presented in the article.

“Frequency of First-line Antibiotic Selection Among US Ambulatory Care Visits for Otitis Media, Sinusitis, and Pharyngitis”