Effect of Antibiotics for Otitis Media on Mastoiditis in Children: A Retrospective Cohort Study Using the United Kingdom General Practice Research Database

1

• Effect of Antibiotics for Otitis Media on
  Mastoiditis in Children A Retrospective Cohort
  Study Using the United Kingdom General Practice
  Research Database

Paula Louise Thompson, Ruth E. Gilbert, Paul F.
Long, Sonia Saxena, Mike Sharland, and Ian Chi
Kei WongPediatrics 2009 123 424-430
Pediatric Residents Journal Club February 5,
2009 Sara Szkola
2
Background

• Otitis media is the main reason children visit
  their doctor in the UK
• Though most cases of otitis media are
  self-limiting, rarely children can go on to
  develop complications such as mastoiditis
• Antibiotic prescriptions for otitis media have
  decreased in recent years, and other studies have
  documented a concurrent increase in hospital
  admissions (1, 2, 3)
• This study aimed to determine if this is a causal
  relationship between decreased antibiotic use and
  increased complications (ie, mastoiditis), or if
  it is due to changes in access to services,
  referral patterns and coding of routine data
• That is, does it make sense to give antibiotics
  for otitis media for the purpose of preventing
  mastoiditis?

3
Study Design

• Data were drawn from the General Practice
  Research Database, a database that draws from
  practices all over the UK and includes
  information on all active patients
• Children aged 3 months to 15 years who were in
  this database from 1990-2006 were included
• 2 exclusions children who werent with up to
  standard practice (MD not recording all data),
  and temporarily registered children (ie, 1
  visit to different MD than their own)

4
Study Design Continued
5
Data Analysis

• Age specific incidence rates (along with
  calendar-year specific incidence rates) for
  mastoiditis with and without otitis media were
  calculated per 1000 child years at risk
• Incidence rates were also calculated for otitis
  media and for antibiotic-prescribing for otitis
  media
• Then they calculated the relative risk of
  children with otitis media developing mastoiditis
  with and without antibiotc treatment

6
What are all these statistics?

• Incidence rate per 1000 children years means the
  number of new cases per unit of person-time at
  risk. That is.
• If the incident rate is 10 per 1000 children
  years, then in a population of 1000 children, 10
  would be expected to develop otitis media in 1
  year. If you observed the population for 2 years
  instead of 1, then 20 would be expected to
  develop otitis media (you divide the number of
  cases by the number of years).

7
What did they find out?

• Of the approximately 2.6 million children in the
  database, 854 had mastoiditis. Only about 1/3
  (288 children) of this group had preceding otitis
  media.
• There were approximately 1.2 million episodes of
  otitis media of these children, 288 went on to
  develop mastoiditis.
• For the children with otitis media, antibiotic
  treatment lowered the risk of mastoiditis from
  3.8/10,000 to 1.8 per 10,000.

8
More interesting trends

• They also found that while the incidence of
  antibiotic prescribing for otitis media steadily
  decreased over the time period of the study
  (1990-2006), the incidence of mastoiditis
  remained stable.

9
What does this mean for me?

• You would have to treat an average of 4831
  episodes of otitis media to prevent one episode
  of mastoiditis
• Decreased prescribing of antibiotics did not
  cause an increase in mastoiditis
• Antibiotics should not be routinely given for
  otitis media for the purpose of preventing
  mastoiditis

10
Critical Review
11
Was the type of study strong?

• Yes
• This article contained a combination of cohort
  (otitis media group) and case control
  (mastoiditis group) studies, both appropriate to
  look for causation

12
Was the assessment of exposure and outcome free
of bias?

• Mostly.
• The investigators werent determining who had
  mastoiditis or otitis media, and so couldnt
  influence those results
• But, the coding and diagnosis may vary among
  practices (ie, coding for otitis media vs URI)
• Also, not everyone who was prescribed antibiotics
  likely took them

13
Was the association both significant and
clinically important?

• It was statistically signifcant, but
• As the authors clearly mention, not very
  clinically important, because the number needed
  to treat for otitis media to prevent mastoiditis
  is too high to be practical
• While their results wont likely change the trend
  of prescribing less antibiotics for otitis media,
  they can make us feel better that we arent
  putting patients at very high risk for mastoiditis

14
Was the association consistent across studies?

• Yes and No
• Other studies found similar rates of mastoiditis
  in the UK
• However, other studies also pointed to
  mastoiditis peaking in children ages 1-4, while
  this study found the highest incidents in
  adolescents

15
Was the cause shown to precede the effect?

• Yes
• In both the mastoiditis group and the otitis
  media group, otitis media was clearly shown to
  precede the mastoiditis
• In addition, antibiotic administration for otitis
  media preceded mastoiditis

16
Was there a dose-response relationship

• Not applicable to this article, though it would
  have been interesting to investigate incidence of
  mastoiditis with high-dose vs low-dose
  antibiotics, or with severity of otitis media

17
Strengths of this Article

• Large sample size of 2.6 million children
• Long period of time which allowed documentation
  of trend of decreased antibiotic use with stable
  mastoiditis rates

18
Weaknesses of This Article

• Coding practices can vary among practices
• Diagnosis can vary among practices
• Did not define high risk group for which
  antibiotics should be given to prevent mastoiditis

19
Take Home Points

• Helpful article that will not necessarily change
  practice, but
• We can be reassured that in general we are not
  putting children at very high risk for
  mastoiditis if we do not give antibiotics for
  otitis media

20
Acknowledgments

• Thank you to Dr. Quinlan and Dr. Goldstein for
  their help in this journal club

21
References

• 1) Little P, Watson L, Morgan S, Williamson I.
  Antibiotic prescribing and admissions with major
  suppurative complications of respiratory tract
  infections a data linkage study. Br J Gen Pract.
  2002 52(476)187-193
• 2) Majeed A, Williams S, Jarman B, Aylin P.
  Prescribing of antibiotics and admissions for
  respiratory tract infections in England. Br Med J
  2004 329 (7471) 879.
• 3) Sharland M, Kendall H, Yeates D, et al.
  Changes in antibiotic prescribing in general
  practice and hospital admissions for quinsy,
  mastoiditis and rheumatic fever in children. Br
  Med J. 2005 331 (7512)328-329.