Little P, Stuart B, Andreou P, et al. Primary care randomised controlled trial of a tailored interactive website for the selfmanagement of respiratory infections (Internet Doctor). BMJ Open 2016;6:e009769. doi:10.1136/bmjopen-2015-009769

There is a labeling mistake in the analysis of this paper. Although the mistake alters the estimates very slightly it does not alter the inferences.

The misunderstanding was that the 24 week follow-up (the last data point) was thought to be 20 weeks, and the real 20 week data was then not included. The most complete use of the data is therefore to include both the 20 and 24 weeks data. If the complete 24 weeks of data is used throughout this means the following changes to the tables: In the Abstract, the Results section should read: “Results 3044 participants were recruited. 852 in the intervention group and 920 in the control group reported one or more RTIs, among whom there a modest increase in NHS Direct contacts in the intervention group (intervention 44/1734 (2.5%) versus control 24/1842 (1.3%); multivariate Risk Ratio (RR) 2.53 (95% CI 1.10 to 5.82, p=0.029)). Conversely reduced contact with doctors occurred (283/1734 (16.3%) vs 368/1845 (20.0%); risk ratio 0.71, 0.53 to 0.95, p=0.019). Reduction in contacts occurred despite slightly longer illness duration (11.3 days versus 10.9 days respectively; multivariate estimate 0.48 days longer (-0.16 to 1.12, p=0.141) and more days of illness rated moderately bad or worse illness (0.53 days; 0.12 to 0.94, p=0.012). The estimate of slower symptom resolution in the intervention group was attenuated when controlling for whether individuals had used webpages which advocated ibuprofen use (length of illness 0.22 days, −0.51 to 0.95, p=0.551; moderately bad or worse symptoms 0.36 days, −0.08 to 0.80, p=0.105). There was no evidence of increased hospitalisations (risk ratio 0.13; 0.02 to 1.01; p=0.051).”

Table 2 is the characteristics of those who reported at least one illness and that does not change. The main impact is on the numbers in Table 3 (monthly report) and Table 4 (notes reviews at 20 weeks in those who reported an infection). These now use 24 weeks of data. There were no changes to the inferences.

The title of table 5 is changed as it is actually based on 24 weeks rather than 20 weeks of data. Table 6 is revised to use 24 rather than 20 weeks. No change to the inferences. Table 7 is all notes reviews at 12 months and remains unchanged. Table 8 is characteristics of those lost to follow up which also remains unchanged. The relationship with ibuprofen remains unchanged. (length of illness 0.22, −0.51 to 0.95, p=0.551; moderately bad or worse symptoms 0.36, −0.08 to 0.80, p=0.105).

The corrected Tables are shown below: