The Impact of Past Improvements in HRH: The Historical Analysis

[tbhallett]

This PR does the analysis to estimate the impact of health of the increases in Human Resources for Health (healthcare workers) have occurred already in Malawi.

Dominic Nkhoma has provided data on the number of staff (4th July 2024), saved in the team Dropbox.

Basic descriptions of these data are shown below (done in the script src/scripts/data_file_processing/healthsystem/human_resources/analysis_historic_changes_in_hr.py)

[BinglingICL]

Nice trend before 2021. But big drops from 2021 to 2023, overall and consistently across all districts (except Likoma seemly). Was it because Covid-19 relevant or funds declined?

[tbhallett]

Nice trend before 2021. But big drops from 2021 to 2023, overall and consistently across all districts (except Likoma seemly). Was it because Covid-19 relevant or funds declined?

I asked Dominic and he replies,
"""Dear Tim
The 2023 figures are wrong. I will investigate, for now please exclude 2023. Otherwise the rest looks good.

regards

Dominic
"""

So I'll do that for now.

[tbhallett]

Construction of scenarios

[tbhallett]

First run of src/scripts/impact_of_historical_changes_in_hr/scenario_historical_changes_in_hr.py

job_id = historical_changes_in_hr-2024-07-08T145329Z

Figures are

[tbhallett]

So, the analysis and plotting script basically work and give some sensible-seeming results.

Things to do following discussion with team:

• Improve plotting: Show breakdown by cause of deaths/dalys;
• Assemble summary slide deck
• Look at squeeze-factor over time in calibration plots (from Mode 1): Does it show a trend over time consistent with there being more HCW available (i.e., does the squeeze reduce over time) (Not needed: later scale-up scenarios begin in 2020)
• Test whether adjustment to the scaling calculation [to take account of the effect that we might be over-correcting?] make a difference to the 'delta' between the scenarios. (Not needed: later scale-up scenarios begin in 2020)
• Query numbers with Dominic
• Set priority policy to be the HBP (?)
• Re-run when confirmed analysis design per the abvove

[tbhallett]

START AGAIN -- new data file provided, showing more modest changes. Will re-do analysis,

[tbhallett]

With new email from Dominic, giving final scale-up pattern.... new graphs below and job-id is:

FULL-SIZE RUN: historical_changes_in_hr-2024-08-14T090007Z

MINI-RUN (BACKUP): historical_changes_in_hr-2024-08-14T174730Z

"SUPPLEMENTARY RESULTS" RUN (pop 20_000, EHPIII and max hcs & hs func, to take advantage of incr resources): historical_changes_in_hr-2024-08-15T094244Z

[BinglingICL]

With new email from Dominic, giving final scale-up pattern.... new graphs below and job-id is:

FULL-SIZE RUN: historical_changes_in_hr-2024-08-14T090007Z

MINI-RUN (BACKUP): historical_changes_in_hr-2024-08-14T174730Z

RUN (pop 20_000, EHPIII and max hcs & hs func, to take advantage of incr resources): historical_changes_in_hr-2024-08-15T094244Z

Thanks for keep sharing the latest results. It interests me a lot as I see the hcw staff count decreased a little bit but not much in the year of 2020, when the pandemic happened. And then from 2021, the hcw staff continue to expand, which shows some resilience of the healthcare system I would suppose.

[tbhallett]

Thanks for keep sharing the latest results. It interests me a lot as I see the hcw staff count decreased a little bit but not much in the year of 2020, when the pandemic happened. And then from 2021, the hcw staff continue to expand, which shows some resilience of the healthcare system I would suppose.

Thanks: yes, maybe that is a part of it. I think it also shows the reforms that have been in put place starting to have an impact.

[BinglingICL]

Thanks for keep sharing the latest results. It interests me a lot as I see the hcw staff count decreased a little bit but not much in the year of 2020, when the pandemic happened. And then from 2021, the hcw staff continue to expand, which shows some resilience of the healthcare system I would suppose.

Thanks: yes, maybe that is a part of it. I think it also shows the reforms that have been in put place starting to have an impact.

Thanks Tim. Is that a reform about hcw recruitment? By the way, may I know how long does it take to run this scenario?

[BinglingICL]

Thanks for keep sharing the latest results. It interests me a lot as I see the hcw staff count decreased a little bit but not much in the year of 2020, when the pandemic happened. And then from 2021, the hcw staff continue to expand, which shows some resilience of the healthcare system I would suppose.

Thanks: yes, maybe that is a part of it. I think it also shows the reforms that have been in put place starting to have an impact.

Thanks Tim. Is that a reform about hcw recruitment? By the way, may I know how long does it take to run this scenario?

And Tim, I wonder if Dominic also has these years' hcw staff per cadre, especially for Clinical, DCSA, Nursing_and_Midwifery, and Pharmacy?

[tbhallett]

(Noting here that the version of the analysis provided externally is tagged here: https://github.com/UCL/TLOmodel/releases/tag/historical_hrh_analysis_v0.1)