Disitinguishing processes and qualities within NPM constructs
Carl May
Yesterday I contacted co-authors of some of the NPM papers to discuss how the naming of parts of the model was working out in practice and to raise some questions that arose from using the model in practice. Anne MacFarlane has responded to me in a way that opens up a debate about this. I have repeated my email here and Anne will repeat her earlier email to me to the whole newsgroup.
A problem that is evident in the empirical work going on now that uses the model, is about the status of the four constructs/factors. Put simply it is this. These four factors (interactional and skill set workability, and relational and contextual integration) are spoken about in two quite different ways.
(a) They refer to the work through which the normalization of a complex intervention might be accomplished. They qualitatively describe a set of complex and emergent human activities that mediate between a complex intervention and its interpersonal and structural contexts. Normalization is thus explained by reference to social processes - by work invested in contextual integration for example.
(b) They refer to complex interventions, or activities related to them, as things-in-themselves. They define observable or measurable characteristics of a complex intervention and the potential of it to normalize might therefore be regarded as quantifiable - for example, acheiving a normalization process quotient greater than
3.5 (to pull a number out of the air) would do it.
This problem has arisen because the names given to the four factors do imply that they are qualities of a thing rather than elements of a process. We spoke about this today at some length during an Ehealth Implementation Project meeting here at
Newcastle, and we felt we could deal with the problem by giving processes and qualities different names, thus:
|
(Observable) Process |
(Measurable) Quality |
|
Operationalization (by actors of a complex intervention in practice) |
A complex intervention will not be disposed to operationalisation if does not have interactional workability |
|
Mediation (amongst actors of knowledge and practice about a complex intervention) |
A complex intervention will not be disposed to mediation if relational integration is poor |
|
Distribution (amongst actors within division of labour) |
A complex intervention will not be disposed to distribution if it does not have skill set workability |
|
Incorporation (by actors in an institutional or organizational context) |
A complex intervention will not be disposed to contextual integration is poor |
|
|
|
If we can collectively agree this, or something like it, then the model can stabilize itself in practice a bit. But it seems sensible to deal with it before any of the papers currently in process are finalized, and it is important for Tracy, Ben, Jo, Lisa, George and others who are doing field research using the model that this is resolved in a way that is satisfactory in practice as well as for publications already in train. This requires a consensus that it is the right thing to do.
Best, carl
cao...@clinmed.gla.ac.uk
Finally made it! Just to re-iterate my email to you yesterday, as part
of the grooup who discussed this on Tuesday I am supportive of the
name changes that you have suggested. I was struggling to transfer the
conceptual nature of the constructs to empirical data (interview
transcripts) and I certainly feel this will help me, at least.
Look foward to other discussion, once Google finally agrees to let
everyone on ....
Kate
> *(Observable) Process *
>
> *(Measurable) Quality*
linda...@manchester.ac.uk
which has been resubmitted- if that is possible?
> *(Observable) Process *
>
> *(Measurable) Quality*
>
anne.ma...@nuigalway.ie
processes and qualities of the model, May 24th...I did e-mail the
group later on but it hasn't appeared here under this thread so...here
it is again!
Morning Carl, thanks for this. As ever, stimulating to think about
this....
I like the new attention to processes involved in normalization and
the qualities of the 'thing' itself. My understanding of this
distinction would be that the work that e.g. patients and
professionals do to communicate with each other via video conferencing
is a process of work. Their satisfaction/acceptance of this work and
the extent to which it becomes routine for them would indicate the
likelihood that the specific complex intervention has the 'quality' of
high or low interactional workability....Correct?
However, I find I have questions about the terms suggested for process
and I imagine that
(a this is because they have arisen from extensive discussions at your
meeting and are based on shared understanding of those at the meeting
(and I wasn't there!)
(b there is no manual (yet!) to explain and unpack the terms further
to wider audiences
The questions I would ask of these terms are as follows:
'operationalisation' of what? Operationalisation is a broad term that
could potentially refer to operationalisation of the whole of a
planned complex intervention. What is its specific meaning here and
how does it relate to issues of the interaction, communication,
agreements about ease of these issues?
Mediation...guess this would be about mediation of knowledge and
practice about existing workflows and relationships, negotiations
around who has knowledge for what and how it is to be used and so on.
Distribution...okay, this one is easier too, distribution of labour
within a network of actors
Incorporation...the link between this term and my understanding of
contextual integration isn't clear. Is this about incorporation of a
complex intervention into a whole organization? What is the value of
using incorporation rather than 'integration'?
To sum up: the idea of distingushing between processes and qualities
is great. The terms suggested to describe processes seem to me to add
a layer of 'complication' that the model does not need if it is to be
used by wider audiences.
Could the process terms be extended to describe the work that is
undertaken in the process of achieving the 4 qualities?
Could the terminology be taken from the current draft of the BIomed
central paper to ensure consistency and avoid adding more terminology
that is potentially hard for others to decipher?
So e.g. BioMed Central paper Version 009, page 16, Box 3 'Propositions
about endogenous factors of the model' pt 1 re interactional
workability reads 'how does a complex intervention affect the
interaction between patients and professionals and professionals and
professionals?'
Could this be adapted further to the present purpose of describing the
process leading to the quality of interactional workability along the
lines of:
Work undertaken by professionals and patients to interact with each
other in an agreed manner....???
Finally, thanks for setting up the mailing list. It's a great idea.
All the best,
Anne