Param Revamp - Malaria

[mmsuarezcosta]

Files

Files containing parameters:

• parameters.csv: Updated ✅
• MAP_ITNrates.csv: Updated ✅ (Doesn't have param_name, all other cols added)
• MAP_IRSrates.csv : Updated ✅ (Doesn't have param_name, all other cols added)
• WHO_TestData2023.csv - Param names are column headers. I simply added column name 'param_label' and defined as local. Future revision will require transposing the DF. @mnjowe I currently implemented a temp solution in the code. The test_analysis was failing at function is_df_same_size_and_dtype because updated version of this df did not have the param_label. I attempted debugging to determine where the update was happening. I did not find the source, and I have implemented this temp solution:

        # TEMP FIX: Filter out param_label column from WHO_TestData2023 for consistency in df updating
        p['rdt_testing_rates'] = workbook['WHO_TestData2023'].drop(columns=['param_label'], errors='ignore')

• highrisk_districts.csv: Not updated . Kept as-is. Two columns are district (containing district name) and district_num. No prior min/max applicable.
• scaleup_parameters.csv: Updated ✅ Changed target_value to value for consistency. Confirmation required if this change is OK.

Not used in code , recommend deleting if legacy

• PfPR_MAPdata.csv
• inc1000py_MAPdata.csv
• interventions.csv
• WHO_MalReport.csv
• WHOcommodities.csv
  -severe_symptoms.csv

Files continaing calibration data
(including MAP outputs -- uncertain if these values also contain model outputs)

• MAP_CommoditiesData2023
• MAP_InfectionData2023
• mortalityRate_MAPdata
• NMCP
• txCov_MAPdata
• WHO_CaseData2023

Files not in folder ResourceFile_malaria

• Have not been addressed

Local

• temp_max_itn_coverage: Explained here
  . I decided to parameterize this since ITN coverage may vary across contexts, so can have higher/lower limits of what is 'normal'. It’s currently labeled as “temp” because it is a parameter that is used for code efficiency. Ideally this type of parameter could be removed and there are only parameters relating to the health system itself. Currently set to local, but not inherent to local context.
• max_age_incidence_data: The maximum age for which incidence data is available — this is the upper threshold for who is considered "alive" in the module. It may be useful to extrapolate the rates for age 80+ to provide a more complete view of the population. Currently set to local, but not inherent to local context.
• p_sev_anaemia_preg: Probability of severe anaemia in pregnant women with malaria may depend on country-specific baseline anaemia rates
• cfr_severe_malaria: While death rates are often considered universal, the case fatality rate (CFR) is influenced by diagnostic rates (denominator is diagnosed cases) which I believe to be universal. I’ve renamed this variable, so the write-up should be updated accordingly.
• data_end: Refers to the last date for which a given dataset provides data. Currently set to local, but not inherent to local context.

Parameter Nomenclature Formatting

• For parameter names related to DateOffset, to make it more readable, I have included the full name of the HSI that it refers to in the parameter name, including capitalization (e.g., MalariaIPTp_frequency_months). Is this acceptable, or do all parameter names need to be lowercase?
• I have parameterized delay variables. Open discussion as to implement these changes, or keep some hard-coding for readability.
• For inline comments like

# check itn projected values are <=0.7 and rounded to 1dp for matching to incidence tables
I’ve replaced hardcoded values with parameter references for clarity:
# check itn projected values are <= temp_max_itn_coverage and rounded to 1dp for matching to incidence tables
Is this preferred formatting for readability?

Universal

• dur_asym, dur_clin, dur_clin_para: These are based on studies conducted in African contexts, but since they reflect biological or clinical durations, I am labeling as universal

Assumptions

dur_clin: Defined in write-up to be an assumption

Additional Miscellaneous Notes

• Many parameters not mentioned in write-up
• Some parameter names I have updated for clarity (e.g. cfr_severe_malaria) and should be adjusted in the write-up
• Calibration details are not disclosed in the write-up.
• The parameter prob_of_treatment_success is currently set to 1.

[marghe-molaro]

Hi @mmsuarezcosta , I think this too reflects the conversations we've been having up to this point, and I think it's ready for @tdm32 to double check.