Skip to contents

Prepare IPTp Data for MBG Analysis

Source: R/indicator_iptp.R
calc_iptp_mbg.Rd

Prepares cluster-level Intermittent Preventive Treatment in pregnancy (IPTp) data for MBG analysis. Calculates both cumulative (1+, 2+, 3+) and exclusive (exactly 1, exactly 2, exactly 3) dose categories.

Usage

calc_iptp_mbg(
  dhs_ir,
  gps_data,
  indicators = c("iptp_1plus", "iptp_2plus", "iptp_3plus"),
  birth_window_months = 36,
  survey_vars = list(cluster = "v001", interview_date = "v008", birth_date = "b3_01",
    sp_doses = "ml1_1", sp_taken = "m49a_1"),
  gps_vars = list(cluster = "DHSCLUST", lat = "LATNUM", lon = "LONGNUM")
)

Arguments

dhs_ir

DHS Individual Recode (IR) or Children's Recode (KR) dataset.

gps_data

DHS GPS dataset with cluster coordinates.

indicators

Character vector of indicators to calculate:

  • Cumulative:

    • "iptp_1plus": At least 1 dose

    • "iptp_2plus": At least 2 doses

    • "iptp_3plus": At least 3 doses (WHO recommendation)

    • "iptp_4plus": At least 4 doses (requires ml1_1 as sp_doses)

  • Exclusive:

    • "iptp_1only": Exactly 1 dose

    • "iptp_2only": Exactly 2 doses

    • "iptp_3only": Exactly 3 doses

Default: c("iptp_1plus", "iptp_2plus", "iptp_3plus").

birth_window_months

Months to look back for births. Default: 36.

survey_vars

Named list mapping DHS variable names.

gps_vars

Named list for GPS variable mapping.

Value

A list of data.tables (one per indicator).

Details

Supports both IR (Individual Recode) and KR (Children's Recode) formats. The function automatically detects the file type and adjusts variable names.

Methodology: https://github.com/ahadi-analytics/sntmethods/blob/master/inst/methods/iptp_dhs.yml

No ANC attendance restriction is applied; the denominator is all women with a birth in the analysis window and a valid SP response (sp_doses <= 7). The default survey_vars$sp_doses = "ml1_1" is the dose count variable (0-7). If ml1_1 is not available, the helper falls back to sp_taken ("m49a_1", binary 0/1), in which case only IPTp 1+ will produce meaningful results.

See also

calc_iptp_dhs() for survey-weighted estimates

Examples

if (FALSE) { # \dontrun{
iptp_mbg <- calc_iptp_mbg(
  dhs_ir = ir_data,
  gps_data = gps_data,
  indicators = c("iptp_2plus", "iptp_3plus")
)
} # }