Infectious Disease

Topic 1 - Summary
1. Citation:

Conflicting data regarding use of steroids in CAP

2. Summary/Bottom Lines:
  • CAP is one of the leading causes of death worldwide.   In this paper they site a 2005 trial where severe pneumonia patients were given IV hydrocortisone and ended up with better P/F ratios, better overall  CXR, lower CRP, improved MODS scores.  In a 300 pt retrospective study steroid use was associated with decreased mortality.  Two large recent trials showed conflicting results.  One showed steroids shortened hospital stay by 1 day and 1 showed increased recurrence with no change in outcome. 

3. Methods:
  • Double blind, randomized placebo controlled trial involving 7 tertiary care facilities in Switzerland.  There were two groups, one of which received 50 mg prednisone daily for 7 days and the other which got placebo.  The endpoint was “time to clinical stability”.  The analysis was intention to treat analysis. 

  • Inclusion criteria: Age >18, hospital admission with CAP defined as a new infiltrate w/cough, sputum, dyspnea, T of >38, abnormal breath sounds, WBC >10000 or  <4000.

  • Exclusion criteria: active IVDU, acute burns, GI bleeding w/in 3 months, adrenal insufficiency, pregnant or breastfeeding, severe immunosuppression, any condition requiring prednisone of more than 0.5 mg/kg daily, inability to obtain consent

  • Randomization was computer generated to 4 or 6 block size groups.  The placebo and steroids were purchased from the same supplier and were packaged to be identical.  Those involved with the study were blinded to treatment groups.

  • All patients were started on antibiotics “as soon as diagnosis was confirmed.” Most patients treated w/augmentin or ceftriaxone.  If Legionella was suspected or if they required ICU care clarithromycin was added. Study nurses then assessed for “clinical stability” every 12 hrs.   Pt had baseline bloodwork and viral swabs drawn upon admission and then on days 1,3,5 and 7.

  • Screened 2911 pt, 802 randomized, ended up with approx. 360 per treatment group.

  • Primary outcome was time to clinical stability, defined as stable vital signs for at least 24 hrs.  T < 37.8.  HR < 100.  RR <24.  Systolic BP >90 ( not on pressors).  Baseline mental status. SpO2>90% or PaO2 >60 mmHg on RA. PO tolerant.

  • Secondary outcomes: time to discharge, time to earliest possible discharge, recurrence, readmission, ICU admission, mortality, duration of abx use (IV and total), disease activity scores, incidence of complications, side effects.

  • Hypothesis: steroids would reduce time to clinical stability, assumed mortality of 10% for placebo and 7.5% for treatment w/ estimated 75% of patients being clinically stable after 7 days.  From this they calculated that 800 pt would give them a power of 85%.

  • Stats – the patients were analyzed on an intention to treat basis, for primary endpoint an unadjusted hazard ratio and 95% CI with Cox proportional hazards ratio.  Patients who died or never achieved clinical stability were censored from study.  They repeated the analysis on a per protocol basis as a sensitivity analysis as well as did a multivariable Cox analysis to assess for PSI and age as potential confounders. 

  • Results – median time to clinical stability was significantly shorter in the treatment group 3 vs 4.4 days.  hazard ratio of 1.33. Per protocol results were similar. Median time to effective discharge from hospital was shorter in treatment group and duration of IV antibiotic treatment was lower.  No change in all cause mortality, ICU admission, readmission or recurrence.  CRP conc were lower in treatment group and complications of pneumonia were lower in the treatment group.  Adverse events were higher in the treatment group due to increased hyperglycemia requiring insulin treatment but the need for new insulin therapy at 30 days was low in both groups. 

4. Discussion and conclusions
    • Discussion – These findings support original hypothesis that steroids reduce time to clinical stability. Findings are consistent with some earlier studies but contradict the 2010 randomized trial by Snijders which showed no benefit and possible harm.  They mention that their patient population was not as sick as those in prior studies.  This study did not show increased rates of recurrence or readmission.  No effect modification was found from any of the subgroup analysis that were performed. Strengths of study were its size and that they looked at all severity classes of pneumonia.  Limitations were that these were all hospitalized patients, the study was not powered to show a mortality difference and the time to clinical stability endpoint is made up of multiple underlying parameters.  Also hyperglycemia could have led to unblinding.

      Conclusions from JC discussion: Since there is not effect on mortality the attendings present at journal club were not in favor of giving steroids to patients with CAP who did not have any other indication for steroids.   And the number of patients required to give a study enough power to determine any potential mortality benefit is likely so large that is prohibits the study from being done. 

5. Additional Info:



6. Author: Dr. Stephen Howard


Topic 2 - Summary
1. Citation:

Tsoraides SS, Pearl RH, Stanfill AB, et al.  Incision and loop drainage: a minimally invasive technique for subcutaneous abscess management in children. J Pediatric Surgery. 2010 March; 45(3):606-609.

2. Summary/Bottom Lines:

This paper presented a new technique for the management of simple abscess in pediatric surgery patients involving placement of a loop vessel.  Two small incisions on opposite sides of abscess are made, pus is evacuated, cavity irrigated, and then a loop vessel is placed through the incisions and loosely tied above the skin.  They argue this is more practical in the pediatric population given the difficulties of repetitive packing and wound care associated with the classic technique.  They found their technique to have an acceptably low re-operation rate and hypothesize that this technique would also be more cost effective. 

3. Methods: Retrospective study preformed at academic center on all pediatric patients who underwent procedure between Jan-02 & Oct-07.  All procedures performed by 1 of 3 surgeons.  Outcomes: demographic information, duration of symptoms, location of abscess, type of drain used, length of procedure, cultured bacteria, drain duration, antibiotic use, complications.  115 patients total.  5 excluded for pilonidal abscess.  13 lost to follow up.  Report re-operation rate at 6/110, despite 13 lost to follow up.  No other predictors identified.
4. Limits: This paper is limited by it's retrospective nature and the lack of a control.  Additionally, the improper inclusion of the lost to follow up patients in their main data point (re-operation rate) was an oversight.  All of these patients where admitted to the general surgery service, the procedure was performed in the OR, and all most every patient received antibiotics.  None of these features are consistent with our practice in the ED.
5. Additional Info:

My take is that this is mostly a "hey, look what we did" paper.  While the results are interesting, it doesn't prove anything.  A prospective trail with comparison to the standard technique is required.  Dr. Nelson agreed.  Dr. Maldonado mentioned this technique was presented at a conference he attended recently.  While he thinks it is interesting and it might have potential, there isn't enough evidence to convince him to change his practice.

6. Author: Dr. Brandon Charlton