Skip to main content
SearchLoginLogin or Signup

Review 1: "Effect of a Widespread Reduction in Treatment Duration for Group A Streptococcal Pharyngitis on Outcomes and Household Transmission"

Peer reviewers highlighted methodological concerns, including overfitting of logistic models, reliance on non-randomized data, and the absence of a power calculation.

Published onDec 19, 2024
Review 1: "Effect of a Widespread Reduction in Treatment Duration for Group A Streptococcal Pharyngitis on Outcomes and Household Transmission"
1 of 2
key-enterThis Pub is a Review of
Effect of a widespread reduction in treatment duration for group A streptococcal pharyngitis on outcomes and household transmission
Effect of a widespread reduction in treatment duration for group A streptococcal pharyngitis on outcomes and household transmission
Description

Abstract Background the optimal treatment duration for group A streptococcal pharyngitis (GAS-P) is debated. Shorter courses appear inferior for pharyngeal GAS eradication, however the effect of this on household transmission is uncertain. In 2022 a laboratory reporting change drove reduced treatment durations for GAS-P in our region. This study sought to assess the effect of this on outcomes.Methods positive throat swab cultures (TSC) for GAS from two years pre-change until 21 months post-change were matched to antibiotic dispensing data. Logistic models were fitted to examine associations between treatment duration and 30-day repeat antibiotic treatment, repeat GAS-positive TSC, and hospitalisation with complications; 90-day incidence of rheumatic fever; 30-day incident household GAS-P cases.Results 865 patients pre-change and 1604 post-change were included. Pre-change 32.8% received ≤7 days treatment, versus 60.0% post-change (p<0.01). There were no significant differences across any outcome measure at a population level between periods. When the post-change period was examined specifically, no significant differences occurred for any outcome measure for patients receiving five- or seven-days of antibiotics versus ten-days. Patients receiving no antibiotics also had similar outcomes, except for significantly higher odds of 30-day household cases (aOR 2.93, 95%-CI 1.44-5.96, p<0.01).Conclusions shorter treatment durations driven by a change in laboratory reporting resulted in no detectable change in GAS-P outcomes, except for more common household transmission in those receiving no treatment.

RR\ID Evidence Scale rating by reviewer:

  • Potentially informative. The main claims made are not strongly justified by the methods and data, but may yield some insight. The results and conclusions of the study may resemble those from the hypothetical ideal study, but there is substantial room for doubt. Decision-makers should consider this evidence only with a thorough understanding of its weaknesses, alongside other evidence and theory. Decision-makers should not consider this actionable, unless the weaknesses are clearly understood and there is other theory and evidence to further support it.

***************************************

Review: This study aims to evaluate if a high level decision to change the duration of therapy offered to patients with Group A Strep Pharyngitis impacted on complications for the index patient +/- the number of detected secondary infections within the household.

As noted by the authors this is non-randomised observational data which places obvious limitations (missing data, non systematic collection of outcomes etc).

A few specific points:

  1. In the text, point estimates are sometimes shown without range (i.e median duration of abx prescribed but no IQR - I note they are included in tables).

  2. Data appears reliant on 'prescribed course' - probably should mention the limitation that you don't actually know how long people took drugs for.

  3. Are there other facilities/labs that people/their contacts could present to that wouldn't be captured in the system?

  4. I suspect that the various models are probably overfitted - you have at least 6-7 variables in your adjustment set but many of the outcomes you have are <50-60 actual events. Similarly in Table 3 you are showing groups with <10 people in them and then adjusting for multiple other factors. This must be over-fitted.

Comments
1
?
Riba Khan:

Bus Simulator Ultimate offers a surprisingly immersive and realistic experience for fans of bus simulation games. The detailed graphics, varied routes, and realistic driving mechanics make it a standout in its genre. The ability to build your own bus company and manage operations adds an engaging strategic layer. However, the game could improve by optimizing performance on lower-end devices and reducing in-game ads for smoother gameplay. Overall, a great choice for anyone looking to experience the thrill of managing a bus empire!