The authors of this report present the quantitative findings from a large-scale randomised-controlled trial which may be known to some readers as the ProCEED trial (Proactive Care and its Evaluation in Enduring Depression).
A qualitative paper involving interviews with nurses who took part in the trial has already been published (Bennett et al., 2013) and ideally should be read alongside this paper, as it provides useful information to help interpret the quantitative outcomes.
The aim of the study was to see if proactive nurse care helped to reduce either the severity or the number of relapses (or both) experienced by people with recurrent and/or chronic depression over a two year period, in comparison with usual access to a GP.
The trial was designed to compare usual primary care (appointments with a GP) with proactive care appointments with a nurse, scheduled every three months over a two year period for people with chronic and/or recurrent depression. The term ‘proactive care’ was used to describe the treatment arm nurse appointments as these took place even if the service user felt well. Participants in both the treatment arm and the usual care arm could make or cancel appointments with their GP at their own will, according to their degree of distress and symptomatology.
Patients were recruited from 42 general practices in the UK and were eligible for inclusion if they had experienced two or more episodes of depression in the last three years, had a BDI-II scores of 14 or more at baseline (so were at least showing some active depression symptoms at the start of treatment) and reported symptoms fitting chronic or recurrent depression in clinical interview. Having agreed to take part in the trial, participants were aware of being in the treatment or non-treatment arm.
A flow chart of recruitment is provided, showing the processes leading to the 235 controls and 251 treatment cases included in analysis. The two groups were not only demographically similar but were also well matched for numbers of chronic and recurrently depressed individuals, the amount of visits they had made to the GP and practice nurses, and the length of time they had taken antidepressants.
The spread of centres for recruitment gives the trial good generalisability as well as making a large sample size achievable. The use of different locations opened up the possibility of looking at the effect of different provisions of psychological care on the options the proactive care nurses could offer to the participants; this is discussed in more detail in Bennett et al (2013).
The BDI-II was taken at five points during the two year study, whilst the Work and Social Adjustment Scale (WSAS, incorrectly named the Work and Social Activity Scale in the text), a questionnaire of DSM-IV criteria and the EuroQol measure of quality of life were taken at the beginning and end of the trial.
The nurses received three days of additional training, which on the one hand seems pragmatically reasonable if such a model is to be rolled out more widely, but on the other seems somewhat brief if teaching on motivation interviewing is to be delivered, as proved to be the case (see below). The nurses were supervised by telephone every 3-4 months, which again seems minimal but realistic for NHS services.
Despite the ‘proactive’ model seeming to suggest that all appointments would take place in the treatment group, 77 people in this group attended 4 or fewer sessions, and 83 attended 5-8 sessions. This is important information about the acceptability of the approach and in interpreting the results of proactive nurse care relative to the control group. Naturally with an interpersonal approach there will be patients who develop better working relationships with the practice nurse than others, and the importance of the therapeutic relationship for treatment adherence and alleviation of depression symptoms is explored in more depth in Bennett et al (2013).
The BDI-II was the primary measure, and there was not a significantly greater improvement in the treatment group compared to the control group. At baseline, the mean scores were 31.9 (s.d. 9.8) in the intervention group and 33.1 (s.d. 10.6) in the control group; at 24 months, mean scores had fallen in both groups to 22.1 (s.d. 14.2) in the intervention group and 26.0 (s.d. 14.9) in the control group. The rate of change of BDI-II scores over the six measurement points is worth further attention, as both groups seemed to show some plateauing effect, with the intervention group mean score suddenly dropping to 22.1 at 24 months, having been at 25.8 (6 months), 25.2 (12 months) and 25.1 (18 months).
57 individuals in the intervention group and 54 in the control group had had no episodes of depression over the 24 months of the study. 27 intervention and 28 control group members were rated as having chronic depression at 24 months, but 87 intervention group members had recurrent depression compared to 60 control group members. What would lead to this difference is unclear, and as this is not significant it is not discussed in the paper. As there were more people with recurrent depression in the treatment group at baseline (155 compared to 142), it may be that people with chronic depression do better than those with recurrent depression, and it would have been interesting to have a more detailed data set in order to assess possible patterns.
The WSAS score in the intervention group was significantly better than the score in the control group, although only at the p=0.01 level, only by 2.5 points, and with both groups going from scores above 20 to scores between 10 and 20, which are taken as clinically relevant score ranges. Interestingly, the intervention group continued to see their GP just as often as the control group, so the nurse intervention did not reduce GP appointment numbers.
The authors conclude that there is positive evidence for improvements in functioning for people with chronic and/or recurrent depression who receive proactive nurse care as opposed to standard GP contact. They surmise that these improvements may be linked to the nurses’ use of problem-solving approaches. Whilst this might be true, the effects are small and marginally significant, and the lack of differential outcome between groups on the BDI-II might, at first glance, seem a good reason not to pursue the option of utilising nurses as proactive carers for people with chronic depression.
However, that would seem hasty, and interpreting the lack of significant effects detected in this trial requires more thought and nuance. The argument that samples of individuals with chronic and/or recurrent depression are highly heterogeneous and that therefore it is difficult to place much weight on comparing mean scores between two groups is slightly cart-before-horse, given that this is more of a question to ask at the study design stage. Furthermore, such effort was taken to match the two groups that this line of heterogeneity argument seems disingenuous.
Given this, it would have been interesting to have had more of a breakdown of the paths taken by similar individuals within the two groups. What would the scores have looked like had low, medium and high attenders in the intervention group had their scores analysed separately? And if there were differences between these groups, what factors influenced attendance?
Some answers are suggested in the qualitative findings paper (Bennett et al., 2013), namely the therapeutic alliance and, as part of this, the orientation of each of the nurses towards the intervention and the care approach adopted by each nurse.
There are other findings in this paper that might make more sense when seen through the lens of the qualitative findings paper (Bennett et al., 2013). Feedback from the nurses involved suggested that, whilst they already felt competent in problem-solving techniques, they did not feel that the half-day training in motivational interviewing (MI) skilled them up adequately and hence they did not make much use of the MI approach. It is not possible to review the Bennett et al (2013) paper here but suffice to say that if nurses are to be expected to manage the care of patients with major depression with psychological therapy techniques, they perhaps need more training and ongoing consolidation of that training, which in time will prove to be cost-effective.
My sense is that, if the authors had been able to present the quantitative and qualitative data together in a larger paper, a more rounded and more readily interpretable piece of research would be here for discussion; certainly this paper adds to the case made by Greenhalgh and colleagues to the British Medical Journal regarding the value of qualitative research.
What this paper gives us is a large, easily generalisable trial of a feasible primary care approach to helping individuals to manage the considerable mental and functional burdens of major depression. In an NHS increasingly faced with cutbacks and cost-constraints, developing strategies that utilise existing resources is crucial and it is vital that this is done in a scientific rather than reactionary fashion.
Buszewicz M, Griffin M, McMahon EM, Walters K, King M. (2016) Practice nurse-led proactive care for chronic depression in primary care: a randomised controlled trial. Br J Psychiatry. 2016 Apr;208(4):374-80. doi: 10.1192/bjp.bp.114.153312. Epub 2016 Jan 21. [Abstract]
Bennett M, Walters K, Drennan V, Buszewicz M. (2013) Structured pro-active care for chronic depression by practice nurses in primary care: A qualitative evaluation. PLOS ONE, 8(9), e75810.
- Photo credit: Walt Stoneburner via Foter.com / CC BY
- Photo credit: jchapiewsky via Foter.com / CC BY-SA
- Dean Hochman CC BY 2.0