General ID: October 2018

The M. tuberculosis PCR is as efficacious as the sputum acid-fast stain for ruling out active pulmonary TB and decreases length of stay. This prospective cohort study out of UCSF examined 621 patients undergoing evaluation for active pulmonary TB at a single hospital between 2014 and 2016, a period during which the hospital began using the GeneXpert MTB/RIF PCR to guide discontinuation of airborne precautions. Specifically, the hospital instituted a clinical algorithm in which airborne precautions could be stopped in low-risk patients after 2 sputum samples negative for TB by either stain or PCR, and in high risk patients after 3 negative sputums. Among the 301 patients evaluated before the switch to PCR, TB rule-out was completed in only 77%; once the PCR was available, rule-out completion jumped to 98% in cases where the PCR had been ordered.

Here’s the take-home: when performed on two or more samples, the molecular test identified every patient with culture-confirmed TB and excluded TB in every patient with culture-negative sputum. After the change to PCR-based evaluation, the time from first sputum collection to first test result dropped from 18.4 to 4.6 hours, and the time from first sputum collection to final test result dropped from 39.1 to 22.4 hours (p <0.001 for both comparisons). Time spent in airborne isolation decreased from 2.9 to 2.5 days (p=0.001) and median length of stay between the pre and post-implementation periods reduced from 6 to 5 days (p=0.003). As a result, switch to PCR-based TB rule-outs produced an estimated $13,337 savings per patient. PCR-based testing for active TB seems like the way to go for hospital infection control. 30178007

Here’s one more study questioning whether gram-negative bacteremia really needs 10-14 days of therapy. A growing body of evidence suggests that gram-negative bacteremia can be safely treated with short courses of antibiotic therapy, with most studies defining “short” as either seven days or less than 10 days. Here’s one more. The authors reviewed four years’ worth of monomicrobial E.coli bacteremia in adults treated with at least one active drug at a single center (total n=856). The cohort was stratified by receipt of either long (>10 days, median 15 days; n=430) or short (<10 days, median 8 days; n=426) antibiotic courses.

Short-course therapy recipients were older (76 vs 72; p=0.002) and had higher Charlson comorbidity scores (7 vs 6; p=0.02) but less septic shock (4.5% vs 9.8%; p=0.003). In both groups urinary tract infections accounted for about half of the bacteremias, with primary bacteremia, biliary tract infections, and other intraabdominal infections making up most of the remainder. A fifth of the cohort had some form of major immunosuppression (neutropenia, solid organ or stem cell transplantation, corticosteroid use, or uncontrolled HIV infection), and none of these varied in incidence between groups. The short-course recipients were actually less likely to have gotten appropriate empiric antibiotics (73% vs 83%; p=0.001) and trended toward having less source control (24% vs 29%; p=0.08). Otherwise, the groups were similar.

The 90-day mortality rate was 5.5%, and in multivariate analysis was not associated with receipt of a short course of therapy. Relapsed infection occurred in 4.9% of the cohort, and was also not associated with short-course therapy, though it did associate with immunosuppression (HR 4.7; p<0.001) and end-stage liver disease (HR 2.6; p=0.01).

While this study is retrospective, the sample size is decent and the groups are similar. In fact, where the short-course and long-course groups vary, in most cases the variance portended a poorer outcome for the short-course group. So, to see that in fact they still did just as well as those given longer courses of antibiotics really suggests that that second week of antibiotics isn’t adding much to the treatment of E.coli bacteremia. 29371138

The best malaria prophylaxis for your next vacation might be not having to take malaria prophylaxis for your next vacation. The major barrier to the efficacy of travel chemoprophylaxis is that you actually have to take the medication. The authors of this study hypothesized a workaround for travel to malaria-endemic areas: the treatment dose of atovaquone/proguanil, four 250mg/100mg tabs a day for three days, provides protection from malaria for up to 4 weeks, so why not give it prior to travel as chemoprophylaxis?

The study recruited 233 participants from Australian travel medicine clinics who planned a <4-week holiday to a low or medium-risk malaria-endemic area. Participants completed the three days of treatment at least one day prior to departure, and questionnaires were administered to collect data on medication adherence, side effects, and acceptability. Overall, 97.7% of participants took the 3-day course of medication as prescribed, and none of them contracted malaria. While nearly half the patients reported side effects, these were well-tolerated and mostly restricted to days one and two. Participants reported their primary reasons for agreeing to use the experimental prophylaxis were that it was easy to remember (72%), that it required taking fewer tablets (54%), and that it would help scientific research (54% - aww, thanks Australians!). 30281083

OK, maybe uncomplicated Staphylococcus aureus bacteremia (SAB) does exist. One of my ID mentors believes that every case of SAB should be treated with at least four weeks of antibiotics. I’m slowly coming around to this position – especially after reading this recent paper in JAMA, which showed that even among the lowest of the low risk staphylococcal bacteremias, a third of those initially thought to have uncomplicated SAB were later found to have complicated infections. I have yet to see any serious adverse outcomes from the overtreatment of SAB, but have already had many encounters with the devastating effects of SAB undertreatment. But anyway, this prospective cohort study says maybe I’m wrong.

The authors collected data of patients’ first episodes of SAB from at 11 hospitals over 1-3 year periods. They stratified the cohort into high and low risk groups, with high risk being defined by any of the following: persistent bacteremia for more than 7 days, metastatic foci of infection, endocarditis, or the presence of a prosthesis. The primary outcome (“poor outcome”) was a composite of recurrent infection at 30 days and mortality at 90 days.

A total of 1866 cases were analyzed (low risk, n=964; high risk, n=852). As you might expect, the low and high risk groups were very different in baseline characteristics; of these, I think the most important differences were the prevalences of prosthetic devices (0% versus 52%) and metastatic infection (0% versus 37%). In multivariate analysis, age >65yr, pneumonia, higher SOFA score, and chronic liver disease predicted a poor outcome. Administration of less than four weeks of antibiotics in the high risk group versus four weeks or more was associated with poor outcome (aOR 1.7; p=0.05); in contrast, administration of less than two weeks of antibiotics in the low risk group was not significantly associated with poor outcome versus administration of two weeks or more.

So, if you have a SAB with no metastatic source of infection, no prosthesis, and clearance of the bacteremia within a week… you might be able to treat it with less than two weeks of antibiotics? Or hey, at least you can cite this paper in court when you try that and everything goes horribly wrong. Personally, I think I’m going to need RTC data to be comfortable giving less than four weeks of antibiotics in all but the most select and thoroughly worked up cases of SAB. 30287412

Here’s an RTC proving what we’ve been telling women based on anecdote for years: increased daily water intake leads to fewer recurrent UTIs. The authors enrolled 140 healthy premenopausal women with recurrent cystitis (>3 episodes in the past year) who were drinking less than 1.5L of fluid daily and randomized them to either continue their usual fluid intake or drink an additional 1.5L. The primary outcome was frequency of recurrent cystitis in the following year.

The mean age of the group was 36 and they had had a mean 3.3 episodes of recurrent cystitis in the year prior to randomization. During the study period there were 1.7 episodes of cystitis in the water group versus 3.2 episodes in the control group (difference, 1.5 episodes per year; p<0.001). Water recipients also used a mean 1.7 fewer courses of antibiotics and went a mean two months longer between cystitis episodes (p<0.001 for both). Given that this intervention is free, widely available, and very safe (assuming your kidneys, heart and liver are all in more or less working order), counseling patients to drink more water should be a part of every recurrent UTI consult we see. 30285042