In this paper by Hasnain M Dalal and colleagues (BMJ 2010;340:b5631, doi:10.1136/bmj.b5631) the authors have alerted us to errors in two figures and two tables. In figure 10, the column headings in the forest plot and the labelling under the forest plot (Centre based and Home based) should be reversed; this also applies to the version of figure 10 that was published in the abridged article in the print journal. In figure 11 the column headings should also be reversed, except for the data for Carlson (the data points were under the correct headings originally, but the weight should be 4.8% (not 5.8%) and the risk ratio should be 1.21 (1.00 to 1.47) and therefore the data point should lie to the right of the line in the graphic (not the left).The reversed column headings should also omit the prefix “favours,” which was an error in the original manuscript.
In table 6 the values for PCS (physical component score) for Arthur et al and Smith et al at 6 month follow-up have incorrectly been placed in the row for SF-36 instead of the row below. In the same table, the data for MacNew global score at 9 month follow-up for Dalal et al and Taylor et al should be 5.60 (SD 1.12) for home and 5.67 (1.12) for centre based care. In table 9 the mean (SD) number of primary care consultations with centre based care for Jolly et al at 9-12 months should read 0.72 (1.54).
None of these changes affect the conclusions of the paper.
Cite this as: BMJ 2010;340:c1133