Re: PACE trial authors’ reply to letter by Kindlon
Anybody who has physical function of 60 on the sf-36 scale and above means your physical function is equal to 84.1% of a demographically representative English adult population Bowling (1999)(1) (See Figure 1). On first look, this may look like anybody in the PACE trial (2) who has physical function of 60 and above on the SF-36 scale can be classed as recovered because 84.1% of a demographically representative English adult population Bowling (1999) has a score of 60 or above on the SF-36 scale. This is because the PACE trial used the mean (SD) scores from these figures and derived a mean (SD) score of 84 (24) for the whole sample, giving a normal range of 60 or above on SF-36 scale for physical function.
Professor White very kindly sent me the ages of the participants in the PACE trial. The reason I asked for this information was because Bowling (1999) participant sample for physical function was weighted relatively equally across all age groups. When I received the age groups of the PACE trial, I plotted them on a graph against Bowling (1999) and I noticed that the age of the participants in the PACE trial were not a true representation of the Bowling (1999) physical function age ranges.
Figure 2 clearly shows that the PACE trials participants’ ages has a left skewed distribution towards younger people, and the Bowling age groups has a plateau distribution. As younger people have a higher level of physical functioning according to Bowling (1999) (can also be shown by the Figure 2 via the green bar chart), surely the score of 60 on the sf-36 scale and above for normal physical function, should have been higher?
I found that approximately 12% of participants in the PACE trial were over the age of 55, compared with 40% in the Bowling (1999) participants. Surely the PACE trial should have compared each participants’ age to the corresponding physical function in Bowling (1999), in order to work out what a normal range was for physical function? (please note, the age groups given to me by Professor White, didn’t exactly match the Bowling (1999) age brackets, so I had to make some assumptions in which age group the PACE trial participants fitted into, nevertheless, because I used the weighted means I believe you should get a more realistic normal physical function expectations according to the age of participants in the PACE trial).
From my own calculations using the data from Professor White of the age of the participants in the PACE trial, and using this against the corresponding Bowling (1999) age brackets, I found that the mean score of a normal person physical function should have been 90.62 and a corresponding standard deviation of 5.51. Therefore using the PACE trial criteria of 1 stand deviation for someone to meet the criteria of recovery in the SF-36 questionnaire they should be reaching a score of 85.12 for normal physical function (please see my workings in Figure 3). This is significantly higher than 60.
Please note, I am not questioning that the figures show an improvement in physical function, but my concern is the definition of recovery from a score of 60 on the sf-36 scale.
(1)Bowling A, Bond M, Jenkinson C, Lamping DL. Short Form 36 (SF-36) Health Survey questionnaire: which normative data should be used? Comparisons between the norms provided by the Omnibus Survey in Britain, the Health Survey for England and the Oxford Healthy Life Survey. J Public Health Med. 1999 Sep;21(3):255-70. PMID 10528952. http://jpubhealth.oxfordjournals.org/content/21/3/255.full.pdf
(2)White PD, Goldsmith K, Johnson AL, Chalder T, Sharpe M. Recovery from chronic fatigue syndrome after treatments given in the PACE trial. Psychol Med. 2013 Oct;43(10):2227-35. doi: 10.1017/S0033291713000020. PMID: 23363640. http://journals.cambridge.org/action/displayAbstract?fromPage=online&aid...
Competing interests: No competing interests