HIV/TB to do list

[tdm32]

This issue will house the "to-do" list for the HIV and TB modules which could form the basis of full projects or may need to be addressed as further modules are added:

• split out the AIDS-defining cancers
• stratify active TB into extrapulmonary/pulmonary TB
• further work is needed on the links between HIV and malaria, particularly in pregnancy -> poor birth outcomes
• add LAM for TB cases (smear-neg, unable to produce sputum etc)
• import district-level TB data and calibrate mixing parameter
• Consider whether we wish to model the transmission of TB (as perhttps://github.com/UCL/TLOmodel/pull/541)

[tbhallett]

• Examine if (and if so, why) Tb cases grow on an implausibly fast exponential if the consumables for Tb treatment are not available.

[tdm32]

• Examine if (and if so, why) Tb cases grow on an implausibly fast exponential if the consumables for Tb treatment are not available.

This issue no longer applicable with fixed incidence model

[tdm32]

1. put all pregnant F on tld as prep/tx until cease breastfeeding due to high transmission risk in early infection. Most infant transmissions occur during the breastfeeding period, and mainly from incident infections among women who were previously HIV-negative and women living with HIV who drop off antiretroviral treatment not due to failure to diagnose or initiate treatment. Retest during breastfeeding at 9 months (new guidelines), improve retention in mothers, POC viral load at first ANC visit (GeneXpert) -> intensive adherence counselling if high viral load, testing after labour, during breastfeeding plus EID

2. new targets for elimination how much health system would be freed up if elimination targets met, how much personnel hours / consumables freed up. If care is streamlined to 6 month dispensations will this improve adherence? What are the effects on tb, maternal mortality, malaria and everything else

3. diabetes and bp checks at routine hiv appointments - to monitor hypertension risk

4. IPT for all contacts of active TB not just those under 5 years

5. TB - how can testing be more efficient? Optimal strategies to increase testing yield with limited resources 5b. HIV - look at routes into HIV testing - ANC, emergency care, TB diagnosis, newborn testing and check yield to determine most effective for scale-up

6. follow one person with a set of characteristics through their lifetime - what is their likely life-course how does this vary for rural/urban, high wealth/low wealth representative person followed through lifetime, show probabilities of conditions popping up at certain times

7. certain interventions have bigger impact in sub-populations, e.g. NCD will benefit high-income more, childhood infectious diseases interventions will impact lower-income people more look at deaths/DALYs by characteristics to see where the benefits are gained and how interventions could reduce inequalities

8. should HCW be based at more rural facilities even if demand isn't consistently there so when emergencies occur, they can respond

[tdm32]

• look at lifetime HS usage for person infected with HIV at 20: true costs associated with infection include lifetime follow-up appts, costs of treating co-infections / co-morbidities. Consider full lifetime costs when evaluating cost-effectiveness of prevention. Does this vary by sex/age/rural etc.