Bottleneck analysis of antenatal care services

[goodgoodhead]

1. we want to know the correlation between the number of ANC visits and IFA.
2. we want to know the correlation between the number of ANC visits and location, like state, and location.
3. Separate the total number of ANC visits into three groups: none, 1-3, above 4
4. The first ANC visit of what GA weeks, separate before 3 months and after

[ernestguevarra]

1. we want to know the correlation between the number of ANC visits and IFA.
2. we want to know the correlation between the number of ANC visits and location, like state, and location.
3. Separate the total number of ANC visits into three groups: none, 1-3, above 4
4. The first ANC visit of what GA weeks, separate before 3 months and after

Great. Good to know what you are trying to do.

What I just want you to remember is that you should be able to address the task which is asking you to show a bottleneck analysis of antenatal care service coverage. Stated differently, you need to show/identify and explain what the possible reason/s is/are for the levels of antenatal coverage at locality, state and national level.

So, reflecting that to what you want to do, the questions to ask yourselves are:

1. What will the correlation between ANC visits and IFA tell you about the bottlenecks in ANC services?
2. What will the correlation between number of ANC visits and location tell you about bottlenecks in ANC services?

[ernestguevarra]

As we try to get to the finish line for your section of the project, I want us to focus on the questions I've raised above and also to consider some other ways to show our results.

Focus again on the idea of bottlenecks and see what we can report to illustrate this.

For this I think we should get your results organised into data.frame objects either at state level and/or locality level. So maybe something that looks like this:

   state_name      ancGestAge visitsANC  anc4 anyIFA ifa90   vitA postNatal tetanusDoses
 1 Al-Gadarif            3.53      2.99 0.359  0.722 0.449 0.205     0.291         0.655
 2 Al-Gazeera            3.21      3.92 0.521  0.685 0.445 0.169     0.363         0.750
 3 Blue Nile             3.22      3.09 0.392  0.642 0.366 0.198     0.375         0.715
 4 Central Darfur        3.43      1.88 0.235  0.435 0.231 0.162     0.0993        0.369
 5 East Darfur           3.88      2.61 0.338  0.548 0.120 0.154     0.133         0.492
 6 Kassala               3.97      3.68 0.470  0.831 0.576 0.449     0.559         0.558
 7 Khartoum              2.95      5.32 0.777  0.845 0.723 0.375     0.401         0.844
 8 North Darfur          3.88      2.84 0.349  0.659 0.256 0.194     0.352         0.563
 9 North Kourdofan       3.56      4.03 0.557  0.806 0.483 0.193     0.441         0.631
10 Northern              2.66      5.01 0.746  0.903 0.736 0.244     0.568         0.790
11 Red Sea               4.55      2.68 0.286  0.707 0.470 0.361     0.409         0.542
12 River Nile            3.30      3.38 0.473  0.658 0.460 0.0827    0.317         0.696
13 Sinar                 3.26      3.55 0.464  0.773 0.493 0.223     0.184         0.674
14 South Darfur          3.74      2.98 0.409  0.477 0.141 0.0766    0.123         0.515
15 South Kourdofan       3.90      3.79 0.543  0.705 0.284 0.107     0.318         0.737
16 West Darfur           3.37      2.79 0.376  0.563 0.192 0.122     0.186         0.577
17 West Kourdofan        4.05      2.98 0.355  0.654 0.198 0.113     0.199         0.526
18 White Nile            3.48      3.75 0.489  0.818 0.611 0.191     0.417         0.657

[ernestguevarra]

And then using these tables, we can create plots that illustrate bottlenecks. Something like this:

[ernestguevarra]

team @OxfordIHTM/naruto, I haven't seen any new activity in terms of code from you guys so just checking in to see that all is good and that you have a game plan on how to proceed? I will spend more time with your team today so you can let me know.

[ernestguevarra]

Team @OxfordIHTM/naruto, not sure if you are still working on this and the code that is on the naruto branch or the clifford branch has some of the work you did today but I noted that the plot was missing the faceting. I updated the naruto branch and added the facets to the plot. But other than that, I didn't do anything else. Are you still working on this or should I get what you have and your nice plot (and I will add some other final touches to it to make it even nicer) and add it to the report?

The report is currently looking like this: https://oxford-ihtm.io/ihtm-hackathon-2024

and with your input and the plot you made, it will even look nicer!

[ernestguevarra]

Team @OxfordIHTM/naruto, just saw your pull request. Well done! all plots look good.

The report in the link I gave you earlier now has your contributions. You will notice that I added some style inputs into your plots particularly with colour. It is important to use colour appropriately. In this case, colour should be used to convey some other form of grouping or hierarchy in the dataset. You shouldn't add colour just to have colour. In the first three plots, the plots already show results by state. So, to colour each bar a different colour per state looks nice but doesn't really add any layer of clarity to your plot. In some cases, it might even cause confusion as a different colour will tend to mean a different hierarchy or order to the results. So a bit of colour (rather than gray) is nice, but too much can be confusing.

Well done! Good job to each of you and to the team!

[9jagirlinoxford]

Thank you! It was nice that you added some colour, it definitely looks better.