Intended for healthcare professionals

Rapid response to:

Papers

Influence of psychological coping on survival and recurrence in people with cancer: systematic review

BMJ 2002; 325 doi: https://doi.org/10.1136/bmj.325.7372.1066 (Published 09 November 2002) Cite this as: BMJ 2002;325:1066

Rapid Response:

Coping and cancer systematic review: reply

We agree with Watson et al. that there is a fruit and vegetables
issue here, in that are a number of ways of measuring coping in these
studies, and coping itself is not always clearly defined, however a review
which aimed to only include homogenous coping styles would rightly be
accused of taking too narrow a view of coping. We aimed therefore to
include any measure. As regards the importance of
helplessness/hopelessness, we feel that there is little good evidence here
that it has an important influence on outcome. Moreover it also remains to
be shown that interventions to reduce helplessness/hopelessness, or indeed
any other psychological intervention have any impact on survival or
recurrence. However this would be better addressed in a systematic review
of trials.

Mitchell and Kumar suggest that we have missed 13 studies and that
the integrity of a review depends on capturing every known paper,
irrespective of the outcome. This is doubtful. It actually depends on
identifying those papers which meet the inclusion criteria. We had seen
and excluded most of these 13 studies and refer Mitchell and Kumar back to
the methods section of our review. Specifically, we were aiming to include
prospective studies of the association between coping and outcome, and
most of their 13 studies do not fulfill these criteria. For example, one
of the studies which they appear to suggest we should include is an RCT
(Spiegel, 1989) with no data on psychological coping. Most of the other
studies assess psychological variables such as locus of control, social
support, general wellbeing, personality (16PF), Rorschach data, and so on.
However this was clearly not a review of the effects of "any" psychosocial
factor - if it were, we could have included at least 20 other studies
(apart from these 13). Instead, we sought to include only studies of
psychological coping in response to cancer. "Locus of control", for
example, may be an important factor but is not a 'coping style' as defined
by our review. Similarly, social support may be an important factor
influencing outcome, and may well be worth reviewing in its own right,
but, as the title of our paper suggests, this was not a review of social
support. Two studies out of the 13 however may meet the inclusion criteria
- one was an abstract which reports little data or methods (n=117
patients) (Diclemente et al., 1985), and the other a book chapter (n=52
patients) (Wirsching et al.,1988). The latter study appears to use
Grossarth-Maticek's measure of rationality/anti-emotionality - we are
unclear at present whether this could be considered to be a measure of
coping, but in short Mitchell and Kumas's claim that we missed a
significant portion of the literature, and over a thousand patients, is
simply wrong.

Mitchell and Kumar also claim that they cannot understand why we
consider an early assessment of coping style to be a measure of quality.
The 'early measurement of coping style' is a basic criteria for causality
- if a factor is postulated to be causal, logically it has to operate
before the outcome occurs. Measuring coping style close to diagnosis seems
the most appropriate way to assess this factor - the longer after
diagnosis, the greater the possibility that coping style may be a result
of the disease progression (or the experience of treatment) rather than an
effect. This is why we included only prospective studies: because it is
difficult to measure coping in an unbiased fashion after the event.
Similarly we do not know what basis they have for saying that
psychological factors are more likely to be important in 'mild' (sic)
disease than aggressive disease. This seems to be an untested assumption
on their part.

Finally, Watson and several other respondents raise the issue that
there is a need for more good research in the area. This is also echoed by
Cooke who also implies a need in the longer term to keep an eye on
emerging data. We would not disagree with either of these, though trials
of psychological interventions are likely to be particularly useful (even
given some of the respondents criticicisms of RCTs).

Mark Petticrew

Ruth Bell

Duncan Hunter

Competing interests:  
None declared

Competing interests: No competing interests

18 December 2002
Mark P Petticrew
Associate Director
Ruth Bell, Duncan Hunter
MRC SPHSU