 I thank you very much, Stasi. Let me just start my stalker. So, I'll be talking about what does social networks have to do with changing services and good data, everybody. Thanks for joining. So, we've been hearing quite a lot about patient experience this morning, on this afternoon, wherever we are. And one aspect of something that influences patient experience that the patient isn't necessarily aware of is the connections between members of the health care workforce and that they're interacting with, and that can be the case even if the patient is just interacting with a single health care worker, that health care worker is connected to other members of the team. And those connections between health care workers are part of this kind of health system software and this is one way of kind of thinking about the complexity of health systems and being having components of hardware and also software and we know that software aspects of health systems such as ideas and interests, relationships and power and values and norms also influence the quality of care being delivered by health care systems. So when we think about a particular specific health care setting then all of those components will be at play and including aspects of software that will be influencing the quality of care being delivered and that will include things like behavioural norms of staff and also where kind of power and influence lies and therefore how change might happen. We might think about maybe looking at the organogram, this isn't from a health care setting as you can see the organogram that I put here but you know maybe thinking about where the formal authority lies and change of command things like that but then there are also these kind of more day to day informal interactions that happen in life including in in hospitals where people ask others for advice or support with decision making that kind of thing and the two might look quite different. So why are these kind of informal social and social interactions important? Well we know from existing sociological theory that the people that we kind of interact with on a day-to-day basis influence our own behaviours and in hospitals or in health care settings then we know that there are lots and lots of challenges and constraints associated with day-to-day practice so the way in which we kind of find our way and work out the behavioural norms that we adopt are influenced by those around us. Also the people that we spend our time with and our peers influence how our behaviour changes so an example here is from Roger's diffusion of innovation model where people are much more likely to adopt an innovation themselves once their peers those around them have already adopted the innovation. Also the kind of more general structures that we see kind of forming of people's ties with one another can disadvantage certain individuals or groups within that setting and here's a cartoon that's an excellent suggestion Ms Triggs perhaps one of the men here would like to make it. But a challenge with these social interactions that happen all these more informal ties in the workplace one challenge is that they're very difficult to capture, measure or communicate and some very important decision making conversations for example about patient care might even happen in hospital corridors or places that might not be quite so amenable to usual methodologies. So one approach or one option might be to use a methodology called social network analysis which provides a vocabulary and set of measures a quantitative measures for relational data and you can see here in the table below that that data could come from many different sources of evidence for example could be observations or questionnaire interviews or other even other sources but the relational data is analyzed using this network analysis approach and essentially the social network analysis uses graph theory and looks at nodes which are kind of blobs and in our case there would be healthcare workers and edges which is the connection between the two nodes and that edge can either be present or absent it could be non-directional or directional and it could even be weighted and here's an example from the literature and the reference is below and this social network analysis study was conducted on a renaud award in Australia where they asked healthcare workers who they went to for advice about medication and the authors noted that actually there weren't that many there wasn't that much advice seeking going on they felt but where there was advice seeking going on for medication then it was much less likely to happen between individuals from different professional groups so you can see here this is a sociogram which is like a visual representation of social network data and it's a directed sociogram because you can see the arrows here or you might be able to see them they're quite small but you can see that nurses which are in who are blue on this sociogram are kind of grouped to the kind of top end of the sociogram whereas the doctors that are in green are more towards the bottom and there is there is a junior doctor a senior nurse and a pharmacist that are very central to the network overall in this example so with my wonderful colleagues as part of my phd I've been undertaking a realist synthesis looking at the social ties of hospital staff asking the question how wide for whom to what extent and in what context does the social ties of staff within a hospital influence quality of service delivery including quality improvement so we conducted a systematic search and retrieved quite a source of citations from which 75 social network analysis studies were included and these were setting hospitals and we're looking at networks of staff they're mostly from high income settings I won't go into this in detail at all but for those not familiar with realism the realist approach put forth by Paulson and Tilly then essentially one looks for patterns within data to build arguments from which to develop abstracted theory and the unit of analysis in this realist approach is the context mechanism outcome configuration where something about the context some element of context triggers a mechanism which is usually something that's kind of latent and unseen and that causes an outcome to occur so therefore it's a kind of looking for these explanatory pathways so that then one can ideally well after identifying what it is about the context it gives a kind of a target for intervention potentially whereas one might go to manipulate that context in a way that an alternative mechanism might be triggered for a different outcome to occur and again I won't go into this in detail but this was the kind of methodology we used in the analysis to first of all configure the CMOCs and then to develop the program theory to ensure that it was supported by the data and refined and the program theory that resulted from this work consisted of four domains organising domains which were emergent from the data and included 35 context mechanism outcome configurations and these domains were social groups hierarchy bridging distance and discourse and I'll just give you an example of social groups there were 13 CMOCs in this domain and a summary of the domain was that hospital staff prefer to communicate with colleagues who are similar to themselves and with whom they share trust however this can create boundaries silos and redundant information within pockets of the workforce and different behavioural norms adopted by members of different groups dominated by individual by influential individuals and just an example of a context mechanism outcome configuration from this domain when a healthcare worker has an existing reciprocal relationship with a peer that's the context they preferentially seek advice or support from that person which is the outcome because they trust the person mechanism and feel comfortable so the conclusions that we found from this synthesis were that individual healthcare workers are subject to their social position in the workforce at large determining both their access to information and support from others and the possible actions available to them and potential targets of intervention are identified for improving for improving communication and distribution of influence and power and thereby supporting behavioural change and quality improvement initiatives in hospitals so going on from from the realist synthesis then I'm now working with colleagues on the pathway study which we're hoping will start early next year and it's being led by Conrad Wanyama and again this pathway study is looking at these interactions within the healthcare team and we'll be using a realist evaluation approach and to look at ties between staff involved in the delivery of care to newborns in two hospitals in Nairobi and to collect the data and for that realist evaluation we'll be using non-participant observation and in-depth interviews and social network analysis and then we'll use that diverse data to inform a co-design event with stakeholders and the objective I suppose of the work will be to try to find a way of better understanding these relational ties and the influence they have on quality of care and then to try to identify ways in which we might usefully use that understanding to improve quality through and this is part of a of I'm sure many of us are aware of the CIN initiative so this is part of a much wider programme of work I'm looking to intervene to improve quality of newborn care in Kenya and so thank you very much questions very welcome