Duration of symptoms of respiratory tract infections in children: systematic review

The authors would like to thank Tähtinen and Ruohola for their interest in our review and their careful scrutiny of the results tables (1). In particular we appreciate the additional information that they clarified from their own study (2). This study presented Kaplan-Meier curves including the time to treatment failure (Tähtinen et al Figure 2) and the time to resolution of ear pain as reported by parents (Tähtinen et al, Supplement Figure 2) each of which had different denominators (160 and 80 respectively). In addition as they noted, 54 children were removed from the study as they needed rescue medication and the diary used for the outcome only included children who had symptoms recorded in the first 48 hours after recruitment rather than the larger group who had symptoms on study entry. This difference in how denominators were reported was not clear to us when we extracted the data, but based on the new information from this paper, we agree that the proportions of children with earache on days 1 to 8 are: 51%, 32, 27, 10, 7, 1, 1, 0. Extracting data from figures rather than having access to the data is not optimal, so additional data such as this is welcome.

As a result we have recalculated the mean and 95%CI for earache/otitis media in the attached revised Figure 3 (Proportion of children with symptoms of earache). Overall this shows very little change from the original figure, except that it would now be more accurate to state that earache would resolve in 50% of children at 2 to 3 days (rather than 3 days as before) and that it would resolve in 90% of children at 7 to 8 days (same as before).

They also questioned some of the proportions of children with symptoms that we presented, so we have checked these values. In Table 2, the proportion of children with symptoms at follow up for Damoiseaux et al (3) is correct (72% (89/123) at day 4), and for Hoberman et al (4) (47% at day 7), and Mygind et al (5) (38% remained symptomatic at day 3), and Neumark et al (6) (17%, or 15/87 at day 7), are correct in Table 2. The proportion of children with symptoms as presented in Table 2 are therefore correct (with the exception of Tähtinen et al given the new information as noted above). In Table 1 these proportions were allocated to the nearest day in order to increase readability of this table. However, we thank the authors for pointing out that the values for LeSaux et al (7) given in Table 1 the proportion for day 3 should read 22% (53/246) not 78%.

In Table 1 (and similar tables for the other respiratory infections, the values in the column titled ‘No. in sample’ are correct as they are the numbers of participants completing study follow up and match the data in Appendix 2 Characteristics of included studies (column headed ‘Follow up’).

Finally, Tähtinen points out that we used outcomes of both ear pain as well as other relevant symptoms, depending on what was reported in the primary studies, indeed this is clearly outlined as footnotes to the tables. Therefore, the outcome used from Damoiseaux et al was indeed persistent symptoms at day four, assessed by the doctor and defined as persistent earache, fever, crying, or being irritable, while for Hoberman et al, it was a scale (AOM-SOS) which consisted of tugging of ears, crying, irritability, difficulty sleeping, diminished activity, diminished appetite, and fever. The outcome data we used from Neumark et al was indeed the proportion of children who were “back to normal” and was derived from their Figure 1 in that study and correctly reported in our review. The data that Tähtinen suggest we use from Table 2 in this study would have been the numbers and proportions of children with “moderate or severe pain” as the data for “some pain” are not included. In general, as Figure 3 shows, the studies which used these combinations of symptoms as outcomes typically reported longer duration than the studies that used solely ear pain. Of course the key information for parents with a child whose predominant symptom complex is indeed ear ache (presumably due to AOM), is when their child will be better overall, with normal activity, sleeping, eating etc. Therefore we feel the information on the duration of symptom cluster in which ear ache predominates that we present is most useful for parents.

Sincerely,

Matthew Thompson
Peter Blair
Carl Heneghan
David Buckley
Alastair Hay

References
1)Miscalculations in the article “Duration of symptoms of respiratory tract infections in children: systematic review: Tähtinen PA, Ruohola A. BMJ Rapid Response 21 December 2013
2)Tähtinen PA, Laine MK, Huovinen P, Jalava J, Ruuskanen O, Ruohola A. A placebo-controlled trial of antimicrobial treatment for acute otitis media. N Engl J Med 2011;364:116-26
3)Damoiseaux RA, van Balen FA, Hoes AW, Verheij TJ, de Melker RA. Primary care based randomised, double blind trial of amoxicillin versus placebo for acute otitis media in children aged under 2 years. BMJ 2000;320:350-4.
4)Hoberman A, Paradise JL, Rockette HE, Shaikh N, Wald ER, Kearney DH, et al. Treatment of acute otitis media in children under 2 years of age. N Engl J Med 2011;364:105-15.
5)Mygind N, Meistrup-Larsen KI, Thomsen J, Thomsen VF, Josefsson K, Sorensen H. Penicillin in acute otitis media: a double-blind placebo-controlled trial. Clin Otolaryngol Allied Sci 1981;6:5-13.
6)Neumark T, Mölstad S, Rosen C, Persson LG, Törngren A, Brudin L, et al. Evaluation of phenoxymethylpenicillin treatment of acute otitis media in children aged 2-16. Scand J Prim Health Care 2007;25:166-71.
7)Le Saux N, Gaboury I, Baird M, Klassen TP, MacCormick J, Blanchard C, et al. A randomized, double-blind, placebo-controlled noninferiority trial of amoxicillin for clinically diagnosed acute otitis media in children 6 months to 5 years of age. CMAJ 2005;172:335-41.