health care seeking at different facility levels

[inesll]

Currently the first appointment at the health systems occurs only at facility level 0 (for non-emergency symptoms) and 1b (for emergency symptoms), through the HSI_GenericEmergencyFirstApptAtFacilityLevel0 and HSI_GenericEmergencyFirstApptAtFacilityLevel1. If sent to another HSI event with another facility_level it counts as 'referral'.

There is no first appointments happening at level 1a (heath centres) or 2 (district hospitals) - which happens in practice

Care seeking module should be able to assertain the level at which patients seek care (level 0, 1a, 1b, 2 outpatient) The DHS 2015-16 dataset has information on care seeking at facility levels for:

• Women and men seeking treatment for STIs
• HIV/AIDS
• Child health
• other heath issues
• Delivery in health facility/home + Skilled Assistance during Delivery

Care seeking facilities options in the questionnaire survey:

PUBLIC SECTOR PUBLIC SECTOR GOVERNMENT HOSPITAL . GOVERNMENT HEALTH CENTER GOVERNMENT HEALTH POST/OUTREACH MOBILE CLINIC HSA OTHER PUBLIC SECTOR

CHAM/MISSION HOSPITAL HEALTH CENTER PRIVATE MEDICAL SECTOR PRIVATE HOSPITAL CLINIC PHARMACY PRIVATE DOCTOR MOBILE CLINIC HSA OTHER PRIVATE MEDICAL SECTOR

BLM MACRO YOUTH DROP IN CENTRE

OTHER SOURCE SHOP TRADITIONAL PRACTITIONER MARKET ITINERANT DRUG SELLER OTHER

[tbhallett]

Thanks @inesll Great idea for an elaboration to the healthcare seeking module.

[tbhallett]

Thanks again for this issue @inesll

We now have a student -- Shiqi Su --- at Imperial who is going to work on this, supervised by @BinglingICL

[inesll]

Oh that is fantastic!!! :) May be worth checking with Joe, he co-supervised a MSc student on this idea for their project - @joehcollins ?

[joehcollins]

Thanks @inesll - my student only looked at care seeking for children with cough +/- fever so something more comprehensive is probably needed!!

[tbhallett]

Thanks @inesll and @joehcollins -- I'm sure we'll come and ask for your advice and help during this project.

[BinglingICL]

By reading the IHS 2019/2020 report (https://microdata.worldbank.org/index.php/catalog/3818), I have summarised the relevant findings about health care seeking as below:

1. For illness or injury (in last 2 weeks/26.9% suffered) 1.1 distribution of healthcare seeking at health facilities by ownership (Table 4-2) Governmental: 50.5%, Private: 5.8%, Mission: 3.2%, Local Pharm/Grocery: 23.1%, Village: 0.9% 1.2 distribution of diagnosis health facilities by types (To summarise) Hospital (1b/2/3), Other health facility (1a), HSA (0)

2. For chronic illness (8.7% suffered) 2.1 distribution of diagnosis health facilities by types (Table 4-4) Hospital (1b/2/3): 65.3%, Other health facility (1a): 17.8%, HSA (0): 0.2%

3. For delivery (in last 24 months) 3.1 distribution of delivery places by facility types (Table 4-5) Hospital or Maternity Clinic (1a/1b/2/3): 95.8%, Home: 3.9%, Other: 0.3%

Unfortunately, there is no further division for community/CHAM hospitals, district hospitals and central hospitals in this survey.

Todo: (1) to summarise results for 1.2 using IHS 2019-20 data (HH_MOD_D -> D06) (2) to regenerate Wingston's HSB functions (https://www.tlomodel.org/publications.html) using IHS 2019-20 data and implement it to tlo; importantly, to make symptom-driven non-emergency health care seeking distributed over level 0, level 1a, level 1b and level 2 (currently, only at level 0) (3) to find relevant results from report of Demographic and Heath Survey in Malawi for 2015-16 (https://dhsprogram.com/publications/publication-fr319-dhs-final-reports.cfm) (4) to think about how to deal with health care seeking pattern among levels 1b, 2 and 3...

[joehcollins]

Hi @BinglingICL just to say the DHS does disaggregate facility delivery by health centre and hospital (so level 1a or level 1b/2/3) so at the model I have calibrated to this distinction. However we dont know the split of deliveries by the 'type' of hospital. So at the moment my simplifying assumption is that there is an equal probability of hospital delivery at either level 1b or 2.

[BinglingICL]

Hi @BinglingICL just to say the DHS does disaggregate facility delivery by health centre and hospital (so level 1a or level 1b/2/3) so at the model I have calibrated to this distinction. However we dont know the split of deliveries by the 'type' of hospital. So at the moment my simplifying assumption is that there is an equal probability of hospital delivery at either level 1b or 2.

Hi @joehcollins, thanks very much for the helpful confirmation. Indeed, I am afraid DHS survey does not distinguish between hospitals either. I think your solution is a good one. Will look at our service usage by facility levels and think more. Looking forward to more discussions.

[BinglingICL]

Hi all, please see the progress made in PR #1003.

[tbhallett]

Fixed by #1003