1006. How Should Sore Throats Be Managed? Performance of Four Group A Streptococcal Sore Throat Prediction Rules in New Zealand Children
Session: Poster Abstract Session: Diagnostic Microbiology: Streptococci
Friday, October 9, 2015
Room: Poster Hall
  • IDSA poster FINAL.pdf (3.0 MB)
  • Background: A number of sore throat prediction rules and guidelines exist. Most were developed in first world settings, to reduce antibiotic overuse,  and where co morbidities (resulting in  cough) may be less common. It is not clear whether these rules can safely be exported to high rheumatic fever regions.

    Methods: Data from a randomized controlled school sore throat clinical trial (Lennon et al 2009)  previously performed in  the high  acute rheumatic fever  (ARF) setting of South Auckland, New Zealand (NZ) in  1998-2001 was subjected to further analysis. Children aged 5 – 18 years, who self reported sore throats, were examined by trained lay workers and all had throat swabs taken for culture. There were 12906 episodes of sore throat (804 culture positive and 10857 negative for Group A Streptococcus (GAS).

    Analyses undertaken

    1.Signs and symptoms of GAS positive and negative sore throat encounters were compared

    2.Four GAS sore throat prediction rules (Wald et al 1998, World Health Organization (WHO) 1991, McIsaac revised Centor 1998, and the New Zealand Heart Foundation 2008 (also published in Kerdemelidis et al 2009)) were tested on the data.


    1.No signs and symptoms separately or in combination reliably predicted GAS sore throats.

    2.Prediction rule performance:

    The most sensitive  was the NZ Heart Foundation’s, (54%),

    The highest positive predictive value was 16%, (the McIsaac revised Centor rule),

    The highest specificity was found in the WHO and McIsaac  revised Centor rules (both 99%),

    Negative predictive was 95-96% across all four rules. 


    1.No reliable predictors of GAS sore throat were found for this population at high risk of ARF.

    2. All four prediction rules performed poorly.

    3.Rule and predictor failure may be due to differences in: clinical training of staff, GAS prevalence, comorbidities in the populations (such as asthma, bronchiectesis, and smoking, which may cause cough), and/ or undeclared anti pyretic use.

    4.’First world’ derived guidelines may not necessarily be transferrable to high ARF risk settings.

    Melissa Kerdemelidis, MBChB; MPH;, Planning and Funding, Canterbury District Health Board, Christchurch, New Zealand, Diana R. Lennon, MB, ChB, Population Child and Youth Health, University of Auckland, Auckland, New Zealand and Joanna Stewart, MSc, The University of Auckland, Auckland, New Zealand


    M. Kerdemelidis, None

    D. R. Lennon, None

    J. Stewart, None

    << Previous Abstract | Next Abstract

    Findings in the abstracts are embargoed until 12:01 a.m. PDT, Wednesday Oct. 7th with the exception of research findings presented at the IDWeek press conferences.