Building NHS Data Science Teams.Rmd

---
title: "Building NHS Data Science Teams"
author: "**`r Sys.getenv('MY_NAME')` - `r Sys.getenv('MY_EMAIL')`**<br />`r Sys.getenv('MY_JOBTITLE')`"
institute: "North Central London ICB"
output:
  NCLRtemplates::ncl_presentation: default
---

```{r setup, include = FALSE}
# check that the following environment variables are set
# set these in a file file called .Renviron: e.g.
# MY_NAME=Joe Bloggs
# MY_EMAIL=joe.bloggs@nhs.net
# MY_JOBTITLE=Analyst
stopifnot(Sys.getenv("MY_NAME") != "")
stopifnot(Sys.getenv("MY_EMAIL") != "")
stopifnot(Sys.getenv("MY_JOBTITLE") != "")

library(Cairo)
library(knitr)
library(tidyverse)
library(NCLRtemplates)

# set default options
opts_chunk$set(echo = FALSE,
               fig.width = 7.252,
               fig.height = 4,
               dpi = 300,
               dev.args = list(type = "cairo"))

# Set's the base ggplot theme to the Strategy Unit Theme
theme_set(theme_nclicb())

# uncomment the following lines if you want to use the Strategy Unit theme
# colours by default
# scale_fill_continuous <- partial(scale_fill_ncl, discrete = FALSE)
# scale_fill_discrete <- partial(scale_fill_ncl, discrete = TRUE)
# scale_colour_continuous <- partial(scale_colour_ncl, discrete = FALSE)
# scale_colour_discrete <- partial(scale_colour_ncl, discrete = TRUE)
```

class: ncl-default 

.pull-left[
<br>
<br>
<br>
<br>
<br>
<br>
<br>

# Just my opinions...

### You learn from your ~~mistakes~~ experience

...also, a little about my journey...

]

.pull-right[
<br>
<br>
<br>
<center><img src="https://imgs.xkcd.com/comics/qualifiers.png" alt="XKCD comic, where person is uncertain and using a lot of qualifiers" width="230px"/> <br>
<b>Source:</b> <a href="https://xkcd.com/2158/" style="color: #004DCF">XKCD </a></center>
]
---
class: ncl-default
# My experience

+ Started in public health: cancer registries


+ A year in clinical network: analyst / general techie / dogsbody


+ Acute trust, general analyst & BI developer - UHB


+ Senior analyst role with NHS benchmarking system
  + Significant step in difficulty
  + Consultancy/outward-facing elements
  + Became responsible for statistical elements
  + PhD study, whilst working for UHB


+ Analytical Manager / Lead, Patient Safety - NHSE


+ Deputy Director Specialist Analytics - NCL ICB


+ ICB Head Data Scientist BSOL from November


---
class: ncl-default
# Key thoughts

<br>

### 1. Build allies

### 2. Understand the question

### 3. Be, or employ, a technical leader

### 4. Release your analysts / data scientists / data engineers

### 5. Leadership is a skill - invest in it.

---
class: ncl-default
# 1. Build allies

## Internal
+ Senior sponsorship in you organisation.

+ IT / cyber-security colleagues.

+ Project / service leads.

--

## External
+ Your own local economy.

+ Regional or National Groups - NHS-R.

+ Identify teams/project/people that you like and contact them.

+ More room to work together than you might think.

+ Increasing use of Open Licences too!


---
class: ncl-default
# 2. Understand the question

.pull-left[
+ For each piece of analysis

+ For your team or organisation's priorities

+ Don't be afraid to work at a question until it's clear

+ Use the question as a tool to build your team: e.g. training

+ Don't be afraid to say 'no'

+ People will go around you if you aren't helpful

]


.pull-right[
<br>
<center><img src="https://imgs.xkcd.com/comics/question.png" alt="XKCD comic, saying 'Dear Isaac, do you like me? With possible answers 'yes' and 'no', and extra writing saying 'this is as yet insufficient data for a meaningful answer' " width="300px"/><br>
<b>Source:</b> <a href="https://xkcd.com/1448/" style="color: #004DCF">XKCD </a></center>
]

---
class: ncl-default
# 3. Be, or employ, a technical leader

.pull-left[
<img src="https://res.cloudinary.com/practicaldev/image/fetch/s--maUIxfjX--/c_limit%2Cf_auto%2Cfl_progressive%2Cq_auto%2Cw_880/https://thepracticaldev.s3.amazonaws.com/i/w9iry6gui9ane19djuv5.png" alt="Diagram of technologies used in a tech stack"><p><b>Source:</b><a href="https://dev.to/minchulkim87/my-data-science-tech-stack-2020-1poa" style="color: #004DCF"> My Data Science Tech Stack 2020 </a></p>
]

.pull-right[
+ Get the experience in!

+ Study if you get the opportunity, because you'll learn a lot.

+ You are in a marathon, not a sprint:
 + What tech do you teams currently use?
 + What sort of question do they answer?
 + Do you have any budget?
 + Can you use a question to skill your team up?
 
+ Work out where you want to get to, and how, with a timescale
 
+ Demonstrate return on investment / what-if

+ Ask for forgiveness rather than permission
]

---
class: ncl-default
# 4. Release your analysts / data scientists

+ Recruit good people - you might have to think about how you recruit

+ If you have keen / experienced managers, give them a purpose / direction

+ If you have keen / skilled junior team members

+ Most people don't think they are a data scientist at first

+ Train people, allow them to attend things and grow

+ Allow them the space to try new things


<center><iframe src="https://giphy.com/embed/3o6fIVCYOLGPO1cfYc" width=600" height="240" frameBorder="0" class="giphy-embed" allowFullScreen></iframe><p><a href="https://giphy.com/gifs/david-freedom-braveheart-3o6fIVCYOLGPO1cfYc" style="font-size: 12pt;colour: #004DCF"">via GIPHY</a></p></center>

---
class: ncl-default

# 5. Leadership is a skill - invest in it.

+ As the leader, it's your job to give everyone in your team a direction

+ It's not for you to solve their problems for them

+ Delegate

+ It's ok for people to do things differently than you would

--

+ Vision is important - 
  + Is a good techie a poor leader?
  + Is a good leader a poor techie?

--

+ I don't know many 'born leaders'.  Most of us learn

+ Find people who you respect to learn from

+ This takes time and you will make mistakes

+ Learn ways to counter for your weaknesses

---

.pull-left[
<br><br><br><br>
<br><br><br><br>
# Thanks!
## Questions?

c.mainey1@nhs.net
]

.pull-right[
<br><br><br><br><br>
<center><iframe src="https://giphy.com/embed/MPuTZQqOmYKPK" width="480" height="282" frameBorder="0" class="giphy-embed" allowFullScreen></iframe><p><a href="https://giphy.com/gifs/thegoodfilms-film-the-big-lebowski-dude-MPuTZQqOmYKPK" style="font-size: 12pt;colour: #004DCF">via GIPHY</a></p></center>
]