Package 'mwana'

Title: An Efficient Workflow for Plausibility Checks and Prevalence Analysis of Wasting in R
Description: A simple and streamlined workflow for plausibility checks and prevalence analysis of wasting based on the Standardized Monitoring and Assessment of Relief and Transition (SMART) Methodology <https://smartmethodology.org/>, with application in R.
Authors: Tomás Zaba [aut, cre, cph] , Ernest Guevarra [aut, cph]
Maintainer: Tomás Zaba <[email protected]>
License: GPL (>= 3)
Version: 0.2.1.9000
Built: 2025-03-07 06:21:41 UTC
Source: https://github.com/nutriverse/mwana

Help Index

A sample data of district level SMART surveys with location anonymised

Description

anthro.01 is a two-stage cluster-based survey with probability of selection of clusters proportional to the size of the population. The survey employed the SMART methodology.

Usage

anthro.01

Format

A tibble of 1,191 rows and 11 columns.

Variable Description
area Survey location
dos Survey date
cluster Primary sampling unit
team Enumerator IDs
sex Sex; "m" = boys, "f" = girls
dob Date of birth
age Age in months, typically estimated using local event calendars
weight Weight in kilograms
height Height in centimetres
edema Edema; "n" = no edema, "y" = with edema
muac Mid-upper arm circumference in millimetres

Source

Anonymous

Examples

anthro.01

A sample of an already wrangled survey data

Description

A household budget survey data conducted in Mozambique in 2019/2020, known as IOF (Inquérito ao Orçamento Familiar in Portuguese). IOF is a two-stage cluster-based survey, representative at province level (second administrative level), with probability of the selection of the clusters proportional to the size of the population. Its data collection spans for a period of 12 months.

Usage

anthro.02

Format

A tibble of 2,267 rows and 14 columns.

Variable Description
province The administrative unit level 1 where data was collected
strata Rural or Urban
cluster Primary sampling unit
sex Sex; "m" = boys, "f" = girls
age Calculated age in months with two decimal places
weight Weight in kilograms
height Height in centimetres
edema Edema; "n" = no edema, "y" = with edema
muac Mid-upper arm circumference in millimetres
wtfactor Survey weights
wfhz Weight-for-height z-scores with 3 decimal places
flag_wfhz Flagged WFHZ value. 1 = flagged, 0 = not flagged
mfaz MUAC-for-age z-scores with 3 decimal places
flag_mfaz Flagged MFAZ value. 1 = flagged, 0 = not flagged

Source

Mozambique National Institute of Statistics. The data is publicly available at https://mozdata.ine.gov.mz/index.php/catalog/88#metadata-data_access. Data was wrangled using this package's wranglers. Details about survey design can be read from: https://mozdata.ine.gov.mz/index.php/catalog/88#metadata-sampling

Examples

anthro.02

A sample data of district level SMART surveys conducted in Mozambique

Description

anthro.03 contains survey data of four districts. Each district data set presents distinct data quality scenarios that require a specific prevalence analysis approach. Data from two districts have a problematic WFHZ standard deviation. The data from the remaining two districts are all within range.

This sample data is useful to demonstrate the use of the prevalence functions on a multiple-domain survey data where there can be variations in the rating of acceptability of the standard deviation, hence requiring different analytical approach for each survey domain to ensure accurate estimation.

Usage

anthro.03

Format

A tibble of 943 x 9.

Variable Description
district Survey location
cluster Primary sampling unit
team Survey teams
sex Sex; "m" = boys, "f" = girls
age Calculated age in months with two decimal places
weight Weight in kilograms
height Height in centimetres
edema Edema; "n" = no edema, "y" = with edema
muac Mid-upper arm circumference in millimetres

Source

Anonymous

Examples

anthro.03

A sample data from a community-based sentinel site in an anonymized location

Description

Data was collected from community-based sentinel sites located across three provinces. Each provincial data set presents distinct data quality scenarios, requiring tailored prevalence analysis:

  • Province 1 has a MUAC-for-age z-score standard deviation and age ratio test rating of acceptability falling within range

  • Province 2 has age ratio rated as problematic but with an acceptable standard deviation of MUAC-for-age z-score

  • "Province 3 has both tests rated as problematic

This sample data is useful to demonstrate the use of the prevalence functions on a multiple-domain survey data where variations in the rating of acceptability of the standard deviation exist, hence require different analytical approach for each domain to ensure accurate estimation.

Usage

anthro.04

Format

A tibble of 3,002 x 8.

Variable Description
province Survey location
cluster Primary sampling unit
sex Sex; "m" = boys, "f" = girls
age Calculated age in months with two decimal places
muac Mid-upper arm circumference in millimetres
edema Edema; "n" = no edema, "y" = with edema
mfaz MUAC-for-age z-scores with 3 decimal places
flag_mfaz Flagged MUAC-for-age z-score value; 1 = flagged, 0 = not flagged

Source

Anonymous

Examples

anthro.04

Define wasting

Description

Determine if a given observation in the data set is wasted or not, and its respective form of wasting (global, severe or moderate) on the basis of z-scores of weight-for-height (WFHZ), muac-for-age (MFAZ), raw MUAC values and combined case-definition.

Usage

define_wasting(
  df,
  zscores = NULL,
  muac = NULL,
  edema = NULL,
  .by = c("zscores", "muac", "combined")
)

Arguments

df

A tibble object. It must have been wrangled using this package's wrangling functions for WFHZ or MUAC, or both (for combined) as appropriate.
zscores

A vector of class double of WFHZ or MFAZ values.
muac

An integer or character vector of raw MUAC values in millimeters.
edema

A character vector indicating edema status. Default is NULL. Code values should be "y" for presence and "n" for absence of nutritional edema.
.by

A choice of the criterion by which a case is to be defined. Choose "zscores" for WFHZ or MFAZ, "muac" for raw MUAC and "combined" for combined. Default value is "zscores".

Value

The tibble object df with additional columns named named gam, sam and mam, each of class numeric containing coded values of either 1 (case) and 0 (not a case). If .by = "combined", additional columns are named cgam, csam and cmam.

Examples

## Case-definition by z-scores ----
z <- anthro.02 |>
  define_wasting(
    zscores = wfhz,
    muac = NULL,
    edema = edema,
    .by = "zscores"
  )
head(z)

## Case-definition by MUAC ----
m <- anthro.02 |>
  define_wasting(
    zscores = NULL,
    muac = muac,
    edema = edema,
    .by = "muac"
  )
head(m)

## Case-definition by combined ----
c <- anthro.02 |>
  define_wasting(
    zscores = wfhz,
    muac = muac,
    edema = edema,
    .by = "combined"
  )
head(c)

Identify, flag, and remove outliers

Description

Identify outlier z-scores for weight-for-height (WFHZ) and MUAC-for-age (MFAZ) following the SMART methodology. The function can also be used to detect outliers for height-for-age (HFAZ) and weight-for-age (WFAZ) z-scores following the same approach.

For flagging z-scores, z-scores that deviate substantially from the sample's z-score mean are considered outliers and are unlikely to reflect accurate measurements. For raw MUAC, values that are less than 100 millimeters or greater than 200 millimeters are considered outliers as recommended by Bilukha & Kianian (2023). Including these values in the analysis could compromise the accuracy of the resulting estimates.

To remove outliers, their values are set to NA rather than removing the record from the dataset. This process is also called censoring. By assigning NA values to these outliers, they can be effectively removed during statistical operations with functions that allow for removal of NA values such as mean() for getting the mean value or sd() for getting the standard deviation.

Usage

flag_outliers(x, .from = c("zscores", "raw_muac"))

remove_flags(x, .from = c("zscores", "raw_muac"))

Arguments

x

A numeric vector of WFHZ, MFAZ, HFAZ, WFAZ or raw MUAC values. Raw MUAC values should be in millimetre units.
.from

Either "zscores" or "raw_muac" for type of data to flag outliers from.

Value

An vector of the same length as x of flagged records coded as 1 for a flagged record and 0 for a non-flagged record.

References

Bilukha, O., & Kianian, B. (2023). Considerations for assessment of measurement quality of mid‐upper arm circumference data in anthropometric surveys and mass nutritional screenings conducted in humanitarian and refugee settings. Maternal & Child Nutrition, 19, e13478. Available at https://doi.org/10.1111/mcn.13478

SMART Initiative (2017). Standardized Monitoring and Assessment for Relief and Transition. Manual 2.0. Available at: https://smartmethodology.org.

Examples

## Sample data of raw MUAC values ----
x <- anthro.01$muac

## Apply the function with `.from` set to "raw_muac" ----
m <- flag_outliers(x, .from = "raw_muac")
head(m)

## Sample data of z-scores (be it WFHZ, MFAZ, HFAZ or WFAZ) ----
x <- anthro.02$mfaz

# Apply the function with `.from` set to "zscores" ----
z <- flag_outliers(x, .from = "zscores")
tail(z)

## With `.from` set to "zscores" ----
z <- remove_flags(
  x = wfhz.01$wfhz,
  .from = "zscores"
)

head(z)

## With `.from` set to "raw_muac" ----
m <- remove_flags(
  x = mfaz.01$muac,
  .from = "raw_muac"
)

tail(m)

Calculate child's age in months

Description

Calculate child's age in months based on the date of birth and the date of data collection.

Usage

get_age_months(dos, dob)

Arguments

dos

A Date vector of date of data collection.
dob

A Date vector of the child's date of birth.

Value

A numeric vector of child's age in months. Any value less than 6.0 and greater than or equal to 60.0 months are set to NA.

Examples

## Take two vectors of class "Date" ----
surv_date <- as.Date(
  c(
    "2024-01-05", "2024-01-05", "2024-01-05", "2024-01-08", "2024-01-08",
    "2024-01-08", "2024-01-10", "2024-01-10", "2024-01-10", "2024-01-11"
  )
)
bir_date <- as.Date(
  c(
    "2022-04-04", "2021-05-01", "2023-05-24", "2017-12-12", NA,
    "2020-12-12", "2022-04-04", "2021-05-01", "2023-05-24", "2020-12-12"
  )
)

## Apply the function ----
get_age_months(
  dos = surv_date,
  dob = bir_date
)

A sample mid-upper arm circumference (MUAC) screening data

Description

A sample mid-upper arm circumference (MUAC) screening data

Usage

mfaz.01

Format

A tibble with 661 rows and 4 columns.

Variable Description
sex Sex; "m" = boys, "f" = girls
months Calculated age in months with two decimal places
edema Edema, "n" = no edema, "y" = with edema
muac Mid-upper arm circumference in millimetres

Source

Anonymous

Examples

mfaz.01

A sample SMART survey data with mid-upper arm circumference measurements

Description

A sample SMART survey data with mid-upper arm circumference measurements

Usage

mfaz.02

Format

A tibble with 303 rows and 7 columns.

Variable Description
cluster Primary sampling unit
sex Sex; "m" = boys, "f" = girls
age Calculated age in months with two decimal places
edema Edema, "n" = no edema, "y" = with edema
mfaz MUAC-for-age z-scores with 3 decimal places
flag_mfaz Flagged MUAC-for-age z-score value. 1 = flagged, 0 = not flagged

Source

Anonymous

Examples

mfaz.02

Check whether sample size requirements for IPC Acute Malnutrition (IPC AMN) analysis are met

Description

Data for estimating the prevalence of acute malnutrition used in the IPC AMN can come from different sources: surveys, screenings or community-based surveillance systems. The IPC has set minimum sample size requirements for each source. This function verifies whether these requirements are met.

Usage

mw_check_ipcamn_ssreq(df, cluster, .source = c("survey", "screening", "ssite"))

Arguments

df

A data.frame object to check.
cluster

A vector of class integer or character of unique cluster or screening or sentinel site identifiers. If a character vector, ensure that each unique name represents one location. If cluster is not of class integer or character, an error message will be returned indicating the type of mismatch.
.source

The source of evidence. A choice between "survey" for representative survey data at the area of analysis; "screening" for screening data; "ssite" for community-based sentinel site data. Default value is "survey".

Value

A single row summary tibble with 3 columns containing check results for:

  • n_clusters - the total number of unique clusters or screening or site identifiers;

  • n_obs - the corresponding total number of children in the data set; and,

  • meet_ipc - whether the IPC AMN requirements were met.

References

IPC Global Partners. 2021. Integrated Food Security Phase Classification Technical Manual Version 3.1.Evidence and Standards for Better Food Security and Nutrition Decisions. Rome. Available at: https://www.ipcinfo.org/ipcinfo-website/resources/ipc-manual/en/.

Examples

mw_check_ipcamn_ssreq(
  df = anthro.01,
  cluster = cluster,
  .source = "survey"
)

Estimate the prevalence of combined wasting

Description

Estimate the prevalence of wasting based on the combined case-definition of weight-for-height z-scores (WFHZ), MUAC and/or edema. The function allows users to estimate prevalence in accordance with complex sample design properties such as accounting for survey sample weights when needed or applicable. The quality of the data is first evaluated by calculating and rating the standard deviation of WFHZ and MFAZ and the p-value of the age ratio test. Prevalence is calculated only when all tests are rated as not problematic. If any of the tests rate as problematic, no estimation is done and an NA value is returned. Outliers are detected in both WFHZ and MFAZ datasets based on SMART flagging criteria. Identified outliers are then excluded before prevalence estimation is performed.

Usage

mw_estimate_prevalence_combined(df, wt = NULL, edema = NULL, .by = NULL)

Arguments

df

A tibble object produced by sequential application of the mw_wrangle_wfhz() and mw_wrangle_muac(). Note that MUAC values in df must be in millimeters unit after using mw_wrangle_muac(). Also, df must have a variable called cluster which contains the primary sampling unit identifiers.
wt

A vector of class double of the survey sampling weights. Default is NULL which assumes a self-weighted survey as is the case for a survey sample selected proportional to population size (i.e., SMART survey sample). Otherwise, a weighted analysis is implemented.
edema

A character vector for presence of nutritional edema coded as "y" for presence of nutritional edema and "n" for absence of nutritional edema. Default is NULL.
.by

A character or numeric vector of the geographical areas or identifiers for where the data was collected and for which the analysis should be summarised for.

Details

A concept of combined flags is introduced in this function. Any observation that is flagged for either flag_wfhz or flag_mfaz is flagged under a new variable named cflags added to df. This ensures that all flagged observations from both WFHZ and MFAZ data are excluded from the prevalence analysis.

flag_wfhz flag_mfaz cflags
1 0 1
0 1 1
0 0 0

Value

A summary tibble for the descriptive statistics about combined wasting.

Examples

## When .by and wt are set to NULL ----
mw_estimate_prevalence_combined(
  df = anthro.02,
  wt = NULL,
  edema = edema,
  .by = NULL
)

## When wt is not set to NULL ----
mw_estimate_prevalence_combined(
  df = anthro.02,
  wt = wtfactor,
  edema = edema,
  .by = NULL
)

Estimate the prevalence of wasting based on z-scores of muac-for-age (MFAZ)

Description

Calculate the prevalence estimates of wasting based on z-scores of MUAC-for-age and/or bilateral edema. The function allows users to estimate prevalence in accordance with complex sample design properties such as accounting for survey sample weights when needed or applicable. The quality of the data is first evaluated by calculating and rating the standard deviation of MFAZ. Standard approach to prevalence estimation is calculated only when the standard deviation of MFAZ is rated as not problematic. If the standard deviation is problematic, prevalence is estimated using the PROBIT estimator. Outliers are detected based on SMART flagging criteria. Identified outliers are then excluded before prevalence estimation is performed.

Usage

mw_estimate_prevalence_mfaz(df, wt = NULL, edema = NULL, .by = NULL)

Arguments

df

A data.frame object that has been produced by the mw_wrangle_age() and mw_wrangle_muac() functions. The df should have a variable named cluster for the primary sampling unit identifiers.
wt

A vector of class double of the survey sampling weights. Default is NULL which assumes a self-weighted survey as is the case for a survey sample selected proportional to population size (i.e., SMART survey sample). Otherwise, a weighted analysis is implemented.
edema

A character vector for presence of nutritional edema coded as "y" for presence of nutritional edema and "n" for absence of nutritional edema. Default is NULL.
.by

A character or numeric vector of the geographical areas or identifiers for where the data was collected and for which the analysis should be summarised for.

Value

A summary tibble for the descriptive statistics about wasting.

Examples

## When .by = NULL ----
mw_estimate_prevalence_mfaz(
  df = anthro.04,
  wt = NULL,
  edema = edema,
  .by = NULL
)

## When .by is not set to NULL ----
mw_estimate_prevalence_mfaz(
  df = anthro.04,
  wt = NULL,
  edema = edema,
  .by = province
)

Estimate the prevalence of wasting based on MUAC for survey data

Description

Estimate the prevalence of wasting based on MUAC and/or nutritional edema. The function allows users to estimate prevalence in accordance with complex sample design properties such as accounting for survey sample weights when needed or applicable. The quality of the data is first evaluated by calculating and rating the standard deviation of MFAZ and the p-value of the age ratio test. Prevalence is calculated only when the standard deviation of MFAZ is not problematic. If both standard deviation of MFAZ and p-value of age ratio test is not problematic, straightforward prevalence estimation is performed. If standard deviation of MFAZ is not problematic but p-value of age ratio test is problematic, age-weighting is applied to prevalence estimation to account for the over-representation of younger children in the sample. If standard deviation of MFAZ is problematic, no estimation is done and an NA value is returned. Outliers are detected based on SMART flagging criteria for MFAZ. Identified outliers are then excluded before prevalence estimation is performed.

Usage

mw_estimate_prevalence_muac(df, wt = NULL, edema = NULL, .by = NULL)

mw_estimate_smart_age_wt(df, edema = NULL, .by = NULL)

Arguments

df

A tibble object produced by mw_wrangle_muac() and mw_wrangle_age() functions. Note that MUAC values in df must be in millimeters unit after using mw_wrangle_muac(). Also, df must have a variable called cluster which contains the primary sampling unit identifiers.
wt

A vector of class double of the survey sampling weights. Default is NULL which assumes a self-weighted survey as is the case for a survey sample selected proportional to population size (i.e., SMART survey sample). Otherwise, a weighted analysis is implemented.
edema

A character vector for presence of nutritional edema coded as "y" for presence of nutritional edema and "n" for absence of nutritional edema. Default is NULL.
.by

A character or numeric vector of the geographical areas or identifiers for where the data was collected and for which the analysis should be summarised for.

Value

A summary tibble for the descriptive statistics about combined wasting.

References

SMART Initiative (no date). Updated MUAC data collection tool. Available at: https://smartmethodology.org/survey-planning-tools/updated-muac-tool/

See Also

mw_estimate_smart_age_wt() mw_estimate_prevalence_mfaz() mw_estimate_prevalence_screening()

Examples

## When .by = NULL ----
mw_estimate_prevalence_muac(
  df = anthro.04,
  wt = NULL,
  edema = edema,
  .by = NULL
)

## When .by is not set to NULL ----
mw_estimate_prevalence_muac(
  df = anthro.04,
  wt = NULL,
  edema = edema,
  .by = province
)

## An application of `mw_estimate_smart_age_wt()` ----
.data <- anthro.04 |> subset(province == "Province 2")

mw_estimate_smart_age_wt(
  df = .data,
  edema = edema,
  .by = NULL
)

Estimate the prevalence of wasting based on MUAC for non-survey data

Description

It is common to estimate prevalence of wasting from non survey data, such as screenings or any other community-based surveillance systems. In such situations, the analysis usually consists only in estimating the point prevalence and the counts of positive cases, without necessarily estimating the uncertainty. This function serves this use.

The quality of the data is first evaluated by calculating and rating the standard deviation of MFAZ and the p-value of the age ratio test. Prevalence is calculated only when the standard deviation of MFAZ is not problematic. If both standard deviation of MFAZ and p-value of age ratio test is not problematic, straightforward prevalence estimation is performed. If standard deviation of MFAZ is not problematic but p-value of age ratio test is problematic, age-weighting is applied to prevalence estimation to account for the over-representation of younger children in the sample. If standard deviation of MFAZ is problematic, no estimation is done and an NA value is returned. Outliers are detected based on SMART flagging criteria for MFAZ. Identified outliers are then excluded before prevalence estimation is performed.

Usage

mw_estimate_prevalence_screening(df, muac, edema = NULL, .by = NULL)

Arguments

df

A tibble object produced by mw_wrangle_muac() and mw_wrangle_age() functions. Note that MUAC values in df must be in millimeters unit after using mw_wrangle_muac(). Also, df must have a variable called cluster which contains the primary sampling unit identifiers.
muac

A numeric or integer vector of raw MUAC values. The measurement unit of the values should be millimeters.
edema

A character vector for presence of nutritional edema coded as "y" for presence of nutritional edema and "n" for absence of nutritional edema. Default is NULL.
.by

A character or numeric vector of the geographical areas or identifiers for where the data was collected and for which the analysis should be summarised for.

Value

A summary tibble for the descriptive statistics about combined wasting.

References

SMART Initiative (no date). Updated MUAC data collection tool. Available at: https://smartmethodology.org/survey-planning-tools/updated-muac-tool/

See Also

mw_estimate_prevalence_muac() mw_estimate_smart_age_wt()

Examples

mw_estimate_prevalence_screening(
  df = anthro.02,
  muac = muac,
  edema = edema,
  .by = province
)

## With `edema` set to `NULL` ----
mw_estimate_prevalence_screening(
  df = anthro.02,
  muac = muac,
  edema = NULL,
  .by = province
)

## With `.by` set to `NULL` ----
mw_estimate_prevalence_screening(
  df = anthro.02,
  muac = muac,
  edema = NULL,
  .by = NULL
)

Estimate the prevalence of wasting based on weight-for-height z-scores (WFHZ)

Description

Calculate the prevalence estimates of wasting based on z-scores of weight-for-height and/or nutritional edema. The function allows users to estimate prevalence in accordance with complex sample design properties such as accounting for survey sample weights when needed or applicable. The quality of the data is first evaluated by calculating and rating the standard deviation of WFHZ. Standard approach to prevalence estimation is calculated only when the standard deviation of MFAZ is rated as not problematic. If the standard deviation is problematic, prevalence is estimated using the PROBIT estimator. Outliers are detected based on SMART flagging criteria. Identified outliers are then excluded before prevalence estimation is performed.

Usage

mw_estimate_prevalence_wfhz(df, wt = NULL, edema = NULL, .by = NULL)

Arguments

df

A tibble object that has been produced by the mw_wrangle_wfhz() functions. The df should have a variable named cluster for the primary sampling unit identifiers.
wt

A vector of class double of the survey sampling weights. Default is NULL which assumes a self-weighted survey as is the case for a survey sample selected proportional to population size (i.e., SMART survey sample). Otherwise, a weighted analysis is implemented.
edema

A character vector for presence of nutritional edema coded as "y" for presence of nutritional edema and "n" for absence of nutritional edema. Default is NULL.
.by

A character or numeric vector of the geographical areas or identifiers for where the data was collected and for which the analysis should be summarised for.

Value

A summary tibble for the descriptive statistics about wasting.

Examples

## When .by = NULL ----
### Start off by wrangling the data ----
data <- mw_wrangle_wfhz(
  df = anthro.03,
  sex = sex,
  weight = weight,
  height = height,
  .recode_sex = TRUE
)

### Now run the prevalence function ----
mw_estimate_prevalence_wfhz(
  df = data,
  wt = NULL,
  edema = edema,
  .by = NULL
)

## Now when .by is not set to NULL ----
mw_estimate_prevalence_wfhz(
  df = data,
  wt = NULL,
  edema = edema,
  .by = district
)

## When a weighted analysis is needed ----
mw_estimate_prevalence_wfhz(
  df = anthro.02,
  wt = wtfactor,
  edema = edema,
  .by = province
)

Clean and format the output tibble returned from the MUAC-for-age z-score plausibility check

Description

Converts scientific notations to standard notations, rounds off values, and renames columns to meaningful names.

Usage

mw_neat_output_mfaz(df)

Arguments

df

An data.frame object returned by mw_plausibility_check_mfaz() containing the summarized results to be formatted.

Value

A data.frame object of the same length and width as df, with column names and values formatted as appropriate.

Examples

## First wrangle age data ----
data <- mw_wrangle_age(
  df = anthro.01,
  dos = dos,
  dob = dob,
  age = age,
  .decimals = 2
)

## Then wrangle MUAC data ----
data_mfaz <- mw_wrangle_muac(
  df = data,
  sex = sex,
  age = age,
  muac = muac,
  .recode_sex = TRUE,
  .recode_muac = TRUE,
  .to = "cm"
)

## Then run plausibility check ----
pl <- mw_plausibility_check_mfaz(
  df = data_mfaz,
  flags = flag_mfaz,
  sex = sex,
  muac = muac,
  age = age
)

## Now neat the output table ----
mw_neat_output_mfaz(df = pl)

Clean and format the output tibble returned from the MUAC plausibility check

Description

Converts scientific notations to standard notations, rounds off values, and renames columns to meaningful names.

Usage

mw_neat_output_muac(df)

Arguments

df

A tibble object returned by the mw_plausibility_check_muac() function containing the summarized results to be formatted.

Value

A data.frame object of the same length and width as df, with column names and values formatted for clarity and readability.

Examples

## First wrangle MUAC data ----
df_muac <- mw_wrangle_muac(
  df = anthro.01,
  sex = sex,
  muac = muac,
  age = NULL,
  .recode_sex = TRUE,
  .recode_muac = FALSE,
  .to = "none"
)

## Then run the plausibility check ----
pl_muac <- mw_plausibility_check_muac(
  df = df_muac,
  flags = flag_muac,
  sex = sex,
  muac = muac
)

## Neat the output table ----

mw_neat_output_muac(df = pl_muac)

Clean and format the output tibble returned from the WFHZ plausibility check

Description

Converts scientific notations to standard notations, rounds off values, and renames columns to meaningful names.

Usage

mw_neat_output_wfhz(df)

Arguments

df

An tibble object returned by the mw_plausibility_check_wfhz() containing the summarized results to be formatted.

Value

A tibble object of the same length and width as df, with column names and values formatted for clarity and readability.

Examples

## First wrangle age data ----
data <- mw_wrangle_age(
  df = anthro.01,
  dos = dos,
  dob = dob,
  age = age,
  .decimals = 2
)

## Then wrangle WFHZ data ----
data_wfhz <- mw_wrangle_wfhz(
  df = data,
  sex = sex,
  weight = weight,
  height = height,
  .recode_sex = TRUE
)

## Now run the plausibility check ----
pl <- mw_plausibility_check_wfhz(
  df = data_wfhz,
  sex = sex,
  age = age,
  weight = weight,
  height = height,
  flags = flag_wfhz
)

## Now neat the output table ----
mw_neat_output_wfhz(df = pl)

Check the plausibility and acceptability of MUAC-for-age z-score (MFAZ) data

Description

Check the overall plausibility and acceptability of MFAZ data through a structured test suite encompassing checks for sampling and measurement-related biases in the dataset. This test suite follows the recommendation made by Bilukha & Kianian (2023) on the plausibility of constructing a comprehensive plausibility check for MUAC data similar to weight-for-height z-score to evaluate its acceptability when age values are available in the dataset.

The function works on a data.frame returned from wrangling functions for age and for MUAC-for-age z-score data available from this package.

Usage

mw_plausibility_check_mfaz(df, sex, muac, age, flags)

Arguments

df

A data.frame object to check.
sex

A numeric vector for child's sex with 1 = males and 2 = females.
muac

A numeric vector of child's MUAC in centimeters.
age

A vector of class double of child's age in months.
flags

A numeric vector of flagged records.

Details

Whilst the function uses the same checks and criteria as those for weight-for-height z-scores in the SMART plausibility check, the percent of flagged records is evaluated using different cut-off points, with a maximum acceptability of 2.0% as shown below:

Excellent Good Acceptable Problematic
0.0 - 1.0 >1.0 - 1.5 >1.5 - 2.0 >2.0

Value

A single row summary tibble with 17 columns containing the plausibility check results and their respective acceptability ratings.

References

Bilukha, O., & Kianian, B. (2023). Considerations for assessment of measurement quality of mid‐upper arm circumference data in anthropometric surveys and mass nutritional screenings conducted in humanitarian and refugee settings. Maternal & Child Nutrition, 19, e13478. https://doi.org/10.1111/mcn.13478

SMART Initiative (2017). Standardized Monitoring and Assessment for Relief and Transition. Manual 2.0. Available at: https://smartmethodology.org.

See Also

mw_wrangle_age() mw_wrangle_muac() mw_stattest_ageratio() flag_outliers()

Examples

## First wrangle age data ----
data <- mw_wrangle_age(
  df = anthro.01,
  dos = dos,
  dob = dob,
  age = age,
  .decimals = 2
)

## Then wrangle MUAC data ----
data_muac <- mw_wrangle_muac(
  df = data,
  sex = sex,
  age = age,
  muac = muac,
  .recode_sex = TRUE,
  .recode_muac = TRUE,
  .to = "cm"
)

## And finally run plausibility check ----
mw_plausibility_check_mfaz(
  df = data_muac,
  flags = flag_mfaz,
  sex = sex,
  muac = muac,
  age = age
)

Check the plausibility and acceptability of raw MUAC data

Description

Check the overall plausibility and acceptability of raw MUAC data through a structured test suite encompassing checks for sampling and measurement-related biases in the dataset. The test suite in this function follows the recommendation made by Bilukha & Kianian (2023).

Usage

mw_plausibility_check_muac(df, sex, muac, flags)

Arguments

df

A data.frame object to check. It must have been wrangled using the mw_wrangle_muac() function.
sex

A numeric vector for child's sex with 1 = males and 2 = females.
muac

A vector of class double of child's MUAC in centimeters.
flags

A numeric vector of flagged records.

Details

Cut-off points used for the percent of flagged records:

Excellent Good Acceptable Problematic
0.0 - 1.0 >1.0 - 1.5 >1.5 - 2.0 >2.0

Value

A single row summary tibble with 9 columns containing the plausibility check results and their respective acceptability ratings.

References

Bilukha, O., & Kianian, B. (2023). Considerations for assessment of measurement quality of mid‐upper arm circumference data in anthropometric surveys and mass nutritional screenings conducted in humanitarian and refugee settings. Maternal & Child Nutrition, 19, e13478. https://doi.org/10.1111/mcn.13478

SMART Initiative (2017). Standardized Monitoring and Assessment for Relief and Transition. Manual 2.0. Available at: https://smartmethodology.org.

See Also

mw_wrangle_muac() flag_outliers()

Examples

## First wrangle MUAC data ----
df_muac <- mw_wrangle_muac(
  df = anthro.01,
  sex = sex,
  muac = muac,
  age = NULL,
  .recode_sex = TRUE,
  .recode_muac = FALSE,
  .to = "none"
)

## Then run the plausibility check ----
mw_plausibility_check_muac(
  df = df_muac,
  flags = flag_muac,
  sex = sex,
  muac = muac
)

Check the plausibility and acceptability of weight-for-height z-score (WFHZ) data

Description

Check the overall plausibility and acceptability of WFHZ data through a structured test suite encompassing checks for sampling and measurement-related biases in the dataset. The test suite, including the criteria and corresponding rating of acceptability, follows the standards in the SMART plausibility check.

The function works on a data frame returned by this package's wrangling functions for age and for WFHZ data.

Usage

mw_plausibility_check_wfhz(df, sex, age, weight, height, flags)

Arguments

df

A tibble object to check.
sex

A numeric vector for child's sex with 1 = males and 2 = females.
age

A vector of class double of child's age in months.
weight

A vector of class double of child's weight in kilograms.
height

A vector of class double of child's height in centimeters.
flags

A numeric vector of flagged records.

Value

A single row summary tibble with 19 columns for the plausibility check results and their respective acceptability rates.

References

SMART Initiative (2017). Standardized Monitoring and Assessment for Relief and Transition. Manual 2.0. Available at: https://smartmethodology.org.

See Also

mw_plausibility_check_mfaz() mw_plausibility_check_muac() mw_wrangle_age()

Examples

## First wrangle age data ----
data <- mw_wrangle_age(
  df = anthro.01,
  dos = dos,
  dob = dob,
  age = age,
  .decimals = 2
)

## Then wrangle WFHZ data ----
data_wfhz <- mw_wrangle_wfhz(
  df = data,
  sex = sex,
  weight = weight,
  height = height,
  .recode_sex = TRUE
)

## Now run the plausibility check ----
mw_plausibility_check_wfhz(
  df = data_wfhz,
  sex = sex,
  age = age,
  weight = weight,
  height = height,
  flags = flag_wfhz
)

Test for statistical difference between the proportion of children aged 24 to 59 months old over those aged 6 to 23 months old

Description

Calculate the observed age ratio of children aged 24 to 59 months old over those aged 6 to 23 months old and test if there is a statistically significant difference between the observed and the expected.

Usage

mw_stattest_ageratio(age, .expectedP = 0.66)

Arguments

age

A numeric vector of child's age in months.
.expectedP

The expected proportion of children aged 24 to 59 months old over those aged 6 to 23 months old. By default, this is expected to be 0.66.

Details

This function should be used specifically when assessing the quality of MUAC data. For age ratio test of children aged 6 to 29 months old over 30 to 59 months old, as performed in the SMART plausibility check, use nipnTK::ageRatioTest() instead.

Value

A list object with three elements: p for p-value of the difference between the observed and the expected proportion of children aged 24 to 59 months old over those aged 6 to 23 months old, observedR for the observed ratio, and observedP for the observed proportion.

References

SMART Initiative. Updated MUAC data collection tool. Available at: https://smartmethodology.org/survey-planning-tools/updated-muac-tool/

Examples

mw_stattest_ageratio(
  age = anthro.02$age,
  .expectedP = 0.66
)

Wrangle child's age

Description

Wrangle child's age for downstream analysis. This includes calculating age in months based on the date of data collection and the child's date of birth, and setting to NA the age values that are less than 6.0 and greater than or equal to 60.0 months old.

Usage

mw_wrangle_age(df, dos = NULL, dob = NULL, age, .decimals = 2)

Arguments

df

A data.frame object to wrangle age from.
dos

A Date vector of dates when data collection was conducted. Default is NULL.
dob

A Date vector of dates of birth of child. Default is NULL.
age

A numeric vector of child's age in months. In most cases this will be estimated using local event calendars or calculated age in months based on date of data collection and date of birth of child.
.decimals

The number of decimal places to round off age to. Default is 2.

Value

A tibble based on df. The variable age will be automatically filled in each row where age value was missing and both the child's date of birth and the date of data collection are available. Rows where age is less than 6.0 and greater than or equal to 60.0 months old will be set to NA. Additionally, a new variable named age_days of class double for calculated age of child in days is added to df.

Examples

## A sample data ----
df <- data.frame(
  surv_date = as.Date(c(
    "2023-01-01", "2023-01-01", "2023-01-01", "2023-01-01", "2023-01-01"
  )),
  birth_date = as.Date(c(
    "2019-01-01", NA, "2018-03-20", "2019-11-05", "2021-04-25"
  )),
  age = c(NA, 36, NA, NA, NA)
)

## Apply the function ----
mw_wrangle_age(
  df = df,
  dos = surv_date,
  dob = birth_date,
  age = age,
  .decimals = 3
)

Wrangle MUAC data

Description

Calculate z-scores for MUAC-for-age (MFAZ) and identify outliers based on the SMART methodology. When age is not supplied, only outliers are detected from the raw MUAC values. The function only works after age has gone through mw_wrangle_age().

Usage

mw_wrangle_muac(
  df,
  sex,
  muac,
  age = NULL,
  .recode_sex = TRUE,
  .recode_muac = TRUE,
  .to = c("cm", "mm", "none"),
  .decimals = 3
)

Arguments

df

A data.frame object to wrangle data from.
sex

A numeric or character vector of child's sex. Code values should only be 1 or "m" for males and 2 or "f" for females.
muac

A numeric vector of child's age in months.
age

A numeric vector of child's age in months. Default is NULL.
.recode_sex

Logical. Set to TRUE if the values for sex are not coded as 1 (for males) or 2 (for females). Otherwise, set to FALSE (default).
.recode_muac

Logical. Set to TRUE if the values for raw MUAC should be converted to either centimeters or millimeters. Otherwise, set to FALSE (default)
.to

A choice of the measuring unit to convert MUAC values into. Can be "cm" for centimeters, "mm" for millimeters, or "none" to leave as it is.
.decimals

The number of decimal places to use for z-score outputs. Default is 3.

Value

A tibble based on df. If age = NULL, flag_muac variable for detected MUAC outliers based on raw MUAC is added to df. Otherwise, variables named mfaz for child's MFAZ and flag_mfaz for detected outliers based on SMART guidelines are added to df.

References

Bilukha, O., & Kianian, B. (2023). Considerations for assessment of measurement quality of mid‐upper arm circumference data in anthropometric surveys and mass nutritional screenings conducted in humanitarian and refugee settings. Maternal & Child Nutrition, 19, e13478. https://doi.org/10.1111/mcn.13478

SMART Initiative (2017). Standardized Monitoring and Assessment for Relief and Transition. Manual 2.0. Available at: https://smartmethodology.org.

See Also

flag_outliers() remove_flags() mw_wrangle_age()

Examples

## When age is available, wrangle it first before calling the function ----
w <- mw_wrangle_age(
  df = anthro.02,
  dos = NULL,
  dob = NULL,
  age = age,
  .decimals = 2
)

### Then apply the function to wrangle MUAC data ----
mw_wrangle_muac(
  df = w,
  sex = sex,
  age = age,
  muac = muac,
  .recode_sex = TRUE,
  .recode_muac = TRUE,
  .to = "cm",
  .decimals = 3
)

## When age is not available ----
mw_wrangle_muac(
  df = anthro.02,
  sex = sex,
  age = NULL,
  muac = muac,
  .recode_sex = TRUE,
  .recode_muac = TRUE,
  .to = "cm",
  .decimals = 3
)

Wrangle weight-for-height data

Description

Calculate z-scores for weight-for-height (WFHZ) and identify outliers based on the SMART methodology.

Usage

mw_wrangle_wfhz(df, sex, weight, height, .recode_sex = TRUE, .decimals = 3)

Arguments

df

A data.frame object to wrangle data from.
sex

A numeric or character vector of child's sex. Code values should only be 1 or "m" for males and 2 or "f" for females.
weight

A vector of class double of child's weight in kilograms.
height

A vector of class double of child's height in centimeters.
.recode_sex

Logical. Set to TRUE if the values for sex are not coded as 1 (for males) or 2 (for females). Otherwise, set to FALSE (default).
.decimals

The number of decimal places to use for z-score outputs. Default is 3.

Value

A data frame based on df with new variables named wfhz for child's WFHZ and flag_wfhz for detected outliers added.

References

SMART Initiative (2017). Standardized Monitoring and Assessment for Relief and Transition. Manual 2.0. Available at: https://smartmethodology.org.

See Also

flag_outliers() remove_flags()

Examples

mw_wrangle_wfhz(
  df = anthro.01,
  sex = sex,
  weight = weight,
  height = height,
  .recode_sex = TRUE,
  .decimals = 2
)

Convert MUAC values to either centimeters or millimeters

Description

Convert MUAC values to either centimeters or millimeters

Usage

recode_muac(x, .to = c("cm", "mm"))

Arguments

x

A vector of raw MUAC values. The class can either be double or numeric or integer.
.to

Either "cm" (centimeters) or "mm" (millimeters) for the unit of measurement to convert MUAC values to.

Value

A numeric vector of the same length as x with values set to specified unit of measurement.

Examples

## Recode from millimeters to centimeters ----
muac_cm <- recode_muac(
  x = anthro.01$muac,
  .to = "cm"
)
head(muac_cm)

## Using the `muac_cm` object to recode it back to "mm" ----
muac_mm <- recode_muac(
  x = muac_cm,
  .to = "mm"
)
tail(muac_mm)

A sample SMART survey data with weight-for-height z-score standard deviation rated as problematic

Description

A sample SMART survey data with weight-for-height z-score standard deviation rated as problematic

Usage

wfhz.01

Format

A tibble with 303 rows and 6 columns.

Variable Description
cluster Primary sampling unit
sex Sex; "m" = boys, "f" = girls
age Calculated age in months with two decimal places
edema Edema, "n" = no edema, "y" = with edema
wfhz MUAC-for-age z-scores with 3 decimal places
flag_wfhz Flagged weight-for-height z-score value; 1 = flagged, 0 = not flagged

Source

Anonymous

Examples

wfhz.01