skeleton_doc.Rmd

---
title: "HAQ article framework"
output:
  word_document: default
  html_document:
    df_print: paged
---

```{r, include =  FALSE }

knitr::opts_chunk$set(cache = TRUE, echo = FALSE)

library(tidyverse)


```

# Main document

## Introduction

* Amenable mortality defined as ...
* Thought to be a useful comparative measure of effectiveness and quality of healthcare
* Has been updated using GBD methodology and study data 2015
* International comparisons published [@ref]
* First time publication of English sub-national comparison
* Describes
    + Variation
    + Trends
    + Cross comparison with conventional method
    + Models variation with deprivation scores
    + Models variation with other metrics of local healthcare quality and performance 
   


## Methods

* Calculation of HAQ described elsewhere
* In brief (to flesh out):
    + Identifies causes of death using Nolte and McKee list [@ref]
    + Recodes using 'garbage code' methodology
    + Risk adjusts
    + Calculates PCA
    + Constructs index using first 3 componenets
    + Scales index to 0-100
    + Compares index for each country against SDI using frontier analysis
    + Estimates additional gain/loss in index given countries level of socio-economic deprivation
* Obtain pre-calculated HAQ values for each UTLA from GBD
* Calculate summary metrics (tabulate)
* Plot variation and trends per LA via line charts and maps
* Obtain other metrics of healthcare performance and quality ? what
    + ONS area amenable and avoidable mortality rates (model/ correlation)
    + ?? Healthcare spend - ?is this available by LA - ? or estimation from CCGs
    + ?? SHMI - at provider level but map to LA?
    + Other??
* Simple regression model vs UTLA deprivation score and calculate predicted HAQ    

    


## Results

* Tabular summaries
* Maps and charts
* ? Statistical analysis on trends
* Comparison with amenable mortality from ONS
* Context of other metrics
* Model vs deprivation (IMD)
* Estimates of potential improvement




## Discussion

* Novel method
* Literature about vailidty of amenable mortality as a measure of healthcare performance and quality
* What does this add?
* Paucity of indicators of population healthcare quality
* Potential as a monitoring tool?
* Other ideas?


## Conclusions

## Potential figures

```{r}







```


















**********
# Abstract

## Introduction



## Methods

```{r literature_search}

source("pubmedAbstractR.R")

search <- "amenable mortality"

gbd <- pubmedAbstractR(search)

gbd %>%
  group_by(year)%>%
  tally() %>%
  ggplot(aes(year, n)) +
  geom_line() +
  geom_point() +
  labs(tite = paste("Annual articles on ", search))



```



## Results


## Discussion

## Conclusions