Package 'stepssurvey'

Description

A complete analysis pipeline for the WHO STEPwise Approach to NCD Risk Factor Surveillance (STEPS).

Author(s)

Maintainer: Abhijit Pakhare [email protected]

Authors:

• Ankur Joshi

• Lena Charlette

Other contributors:

• WHO STEPS R Pipeline Contributors [contributor]

See Also

Useful links:

• https://github.com/drpakhare/stepssurvey

• Report bugs at https://github.com/drpakhare/stepssurvey/issues

Description

Build all tables from computed results

Usage

build_all_tables(results)

Arguments

Value

A named list of flextable objects. NULL entries excluded.

Description

Creates a horizontal point-and-CI plot (forest plot style) for all key indicators, grouped by domain.

Usage

build_forest_plot(key_indicators, country_name, survey_year)

Arguments

Value

A ggplot2 object.

Description

Creates a radar-style chart showing prevalence of key risk factors on a polar coordinate system for quick visual comparison.

Usage

build_radar_plot(key_indicators, country_name, survey_year)

Arguments

Value

A ggplot2 object.

Description

Generates a list of ggplot2 plots showing key NCD risk factor prevalence with 95% confidence intervals, stratified by sex and age group.

Usage

build_steps_plots(indicators, key_indicators, country_name, survey_year)

Arguments

Details

All plots use the WHO STEPS colour scheme and professional styling. Error bars represent 95% confidence intervals. Prevalence values are displayed on bars/points with light background text.

Value

A named list of ggplot2 objects:

• overview: Horizontal bar chart of key indicators

• tobacco_by_sex: Sex-stratified tobacco use

• bp_by_sex: Sex-stratified blood pressure

• obesity_by_sex: Sex-stratified overweight/obesity

• glucose_by_sex: Sex-stratified blood glucose

• bp_by_age: Age-stratified blood pressure with ribbon CI

• obesity_by_age: Age-stratified overweight/obesity with ribbon CI

• sex_dashboard: Combined 2x2 dashboard of sex-stratified charts (if >=2 sex plots available) NULL entries are preserved in the list.

Examples

test_data <- generate_test_data(n = 500, seed = 42)
  cols <- detect_steps_columns(test_data)
  clean <- clean_steps_data(test_data, cols)
  design <- setup_survey_design(clean)
  all_ind <- compute_all_indicators(design)
  plots <- build_steps_plots(all_ind$results, all_ind$key_indicators, "Test", 2023)
  names(plots)

Description

Generates formatted flextable objects for all available STEPS indicators, with rows for age groups and columns for both sexes combined, males, and females. Tables include 95% confidence intervals.

Usage

build_steps_tables(indicators)

Arguments

Details

Each table has age groups as rows and prevalence (with 95% CI) as a column. The last row shows the total (age-standardised) estimate. Column header styling uses WHO STEPS branding (dark blue background).

Value

A named list of flextable objects, one per indicator. Names correspond to indicators (e.g., current_tobacco, raised_bp). NULL entries are excluded. Prints count of tables generated.

Examples

test_data <- generate_test_data(n = 500, seed = 42)
  cols <- detect_steps_columns(test_data)
  clean <- clean_steps_data(test_data, cols)
  design <- setup_survey_design(clean)
  all_ind <- compute_all_indicators(design)
  tables <- build_steps_tables(all_ind$results)
  names(tables)

Description

Dispatches to the appropriate formatting method based on table type.

Usage

build_table(result)

Arguments

Value

A flextable object, or NULL if the table is not available.

Description

Processes raw STEPS survey data: renames columns, coerces types, derives standard indicators, handles missing values, and applies plausibility checks.

Usage

clean_steps_data(
  data,
  cols,
  age_min = 18,
  age_max = 69,
  bp_sbp_threshold = 140,
  bp_dbp_threshold = 90,
  bmi_overweight = 25,
  bmi_obese = 30,
  glucose_threshold = 7,
  glucose_impaired_threshold = 6.1,
  chol_threshold = 5
)

Arguments

Details

The function performs the following transformations:

• Renames columns to standard names (age, sex, wt_final, etc.)

• Converts numeric strings to appropriate types

• Restricts age to [age_min, age_max]

• Creates WHO standard age groups (18-24, 25-34, etc.)

• Harmonises sex coding to Male/Female

• Derives body mass index (BMI) and categories

• Averages blood pressure readings (last 2 of 3)

• Recodes yes/no variables to logical

• Creates derived risk indicators (raised BP, diabetes, etc.)

• Applies plausibility checks to measurements

• Drops records with missing age or sex

Value

A data frame with standardised and derived variables, ready for survey design setup.

Description

Calculates prevalence of alcohol use from a survey design object. Computes proportions of current alcohol use and heavy episodic drinking, stratified by sex and age group where available.

Usage

compute_alcohol_indicators(design)

Arguments

Value

A named list of survey estimates. Each element contains proportion estimates (as tibble with columns: estimate, lower, upper, etc.) for:

• current_alcohol_total: current alcohol use, overall

• current_alcohol_by_sex: current alcohol use, by sex

• current_alcohol_by_age: current alcohol use, by age group

• heavy_episodic_total: heavy episodic drinking, overall

• heavy_episodic_by_sex: heavy episodic drinking, by sex

• heavy_episodic_by_age: heavy episodic drinking, by age group (if the corresponding variables exist in design)

See Also

compute_all_indicators()

Examples

test_data <- generate_test_data(n = 500, seed = 42)
  cols <- detect_steps_columns(test_data)
  clean <- clean_steps_data(test_data, cols)
  design <- setup_survey_design(clean)
  alcohol_results <- compute_alcohol_indicators(design)

Description

Runs all indicator modules (tobacco, alcohol, diet & physical activity, anthropometry, blood pressure, and biochemical), using the appropriate step-specific survey design for each domain per WHO STEPS methodology:

• Step 1 (behavioural): tobacco, alcohol, diet & physical activity

• Step 2 (physical): anthropometry, blood pressure

• Step 3 (biochemical): biochemical measures

Usage

compute_all_indicators(design)

Arguments

Value

A list with two elements:

• results: a named list containing indicator results grouped by domain (tobacco, alcohol, diet_pa, anthropometry, blood_pressure, biochemical)

• key_indicators: a tibble with columns domain, indicator, estimate, lower, and upper, summarising headline estimates across all domains

Examples

test_data <- generate_test_data(n = 500, seed = 42)
  cols <- detect_steps_columns(test_data)
  clean <- clean_steps_data(test_data, cols)
  design <- setup_survey_design(clean)
  all_indicators <- compute_all_indicators(design)
  names(all_indicators$results)

Description

Iterates through the full steps_table_registry() and computes every table that has available data. Returns a named list of results.

Usage

compute_all_tables(designs, data = NULL)

Arguments

Value

A named list of table results (from compute_table()). Only entries with available == TRUE are included.

Description

Calculates prevalence of overweight, obesity, and central obesity, plus mean BMI and waist circumference, from a survey design object.

Usage

compute_anthropometry_indicators(design)

Arguments

Value

A named list of survey estimates. Each element contains estimates (as tibble with columns: estimate, lower, upper, etc.) for:

• overweight_obese_total: overweight or obese (BMI >=25), overall

• overweight_obese_by_sex: overweight or obese, by sex

• overweight_obese_by_age: overweight or obese, by age group

• obese_total: obese (BMI >=30), overall

• obese_by_sex: obese, by sex

• obese_by_age: obese, by age group

• central_obesity_total: central obesity, overall

• central_obesity_by_sex: central obesity, by sex

• central_obesity_by_age: central obesity, by age group

• bmi_mean_total: mean BMI, overall

• bmi_mean_by_sex: mean BMI, by sex

• waist_cm_mean_total: mean waist circumference, overall

• waist_cm_mean_by_sex: mean waist circumference, by sex (if the corresponding variables exist in design)

See Also

compute_all_indicators()

Examples

test_data <- generate_test_data(n = 500, seed = 42)
  cols <- detect_steps_columns(test_data)
  clean <- clean_steps_data(test_data, cols)
  design <- setup_survey_design(clean)
  anthropometry_results <- compute_anthropometry_indicators(design)

Description

Calculates prevalence of raised glucose, diabetes, impaired glucose tolerance, and raised cholesterol, plus mean fasting glucose and total cholesterol from a survey design object.

Usage

compute_biochemical_indicators(design)

Arguments

Value

A named list of survey estimates. Each element contains estimates (as tibble with columns: estimate, lower, upper, etc.) for:

• raised_glucose_total: raised fasting glucose, overall

• raised_glucose_by_sex: raised fasting glucose, by sex

• raised_glucose_by_age: raised fasting glucose, by age group

• diabetes_total: diabetes, overall

• diabetes_by_sex: diabetes, by sex

• diabetes_by_age: diabetes, by age group

• impaired_glucose_total: impaired fasting glucose, overall

• impaired_glucose_by_sex: impaired fasting glucose, by sex

• impaired_glucose_by_age: impaired fasting glucose, by age group

• raised_chol_total: raised total cholesterol, overall

• raised_chol_by_sex: raised total cholesterol, by sex

• raised_chol_by_age: raised total cholesterol, by age group

• fasting_glucose_mean_total: mean fasting glucose, overall

• fasting_glucose_mean_by_sex: mean fasting glucose, by sex

• total_chol_mean_total: mean total cholesterol, overall

• total_chol_mean_by_sex: mean total cholesterol, by sex (if the corresponding variables exist in design)

See Also

compute_all_indicators()

Examples

test_data <- generate_test_data(n = 500, seed = 42)
  cols <- detect_steps_columns(test_data)
  clean <- clean_steps_data(test_data, cols)
  design <- setup_survey_design(clean)
  biochemical_results <- compute_biochemical_indicators(design)

Description

Calculates prevalence of raised blood pressure and mean systolic and diastolic blood pressure from a survey design object.

Usage

compute_bp_indicators(design)

Arguments

Value

A named list of survey estimates. Each element contains estimates (as tibble with columns: estimate, lower, upper, etc.) for:

• raised_bp_total: raised blood pressure, overall

• raised_bp_by_sex: raised blood pressure, by sex

• raised_bp_by_age: raised blood pressure, by age group

• mean_sbp_mean_total: mean systolic BP, overall

• mean_sbp_mean_by_sex: mean systolic BP, by sex

• mean_sbp_mean_by_age: mean systolic BP, by age group

• mean_dbp_mean_total: mean diastolic BP, overall

• mean_dbp_mean_by_sex: mean diastolic BP, by sex

• mean_dbp_mean_by_age: mean diastolic BP, by age group (if the corresponding variables exist in design)

See Also

compute_all_indicators()

Examples

test_data <- generate_test_data(n = 500, seed = 42)
  cols <- detect_steps_columns(test_data)
  clean <- clean_steps_data(test_data, cols)
  design <- setup_survey_design(clean)
  bp_results <- compute_bp_indicators(design)

Description

Calculates prevalence of insufficient physical activity and low fruit & vegetable intake, plus mean metabolic equivalent (MET) values, from a survey design object.

Usage

compute_diet_pa_indicators(design)

Arguments

Value

A named list of survey estimates. Each element contains estimates (as tibble with columns: estimate, lower, upper, etc.) for:

• insufficient_pa_total: insufficient physical activity, overall

• insufficient_pa_by_sex: insufficient physical activity, by sex

• insufficient_pa_by_age: insufficient physical activity, by age group

• low_fruit_veg_total: low fruit & vegetable intake, overall

• low_fruit_veg_by_sex: low fruit & vegetable intake, by sex

• low_fruit_veg_by_age: low fruit & vegetable intake, by age group

• met_mean_total: mean MET (if available)

• met_mean_by_sex: mean MET by sex (if available) (if the corresponding variables exist in design)

See Also

compute_all_indicators()

Examples

test_data <- generate_test_data(n = 500, seed = 42)
  cols <- detect_steps_columns(test_data)
  clean <- clean_steps_data(test_data, cols)
  design <- setup_survey_design(clean)
  diet_pa_results <- compute_diet_pa_indicators(design)

Description

Takes a table specification from steps_table_registry() and a survey design object, and produces the survey-weighted estimates needed to fill the standard WHO STEPS data book table.

Description

This is the main workhorse: given one registry entry and a survey design, it dispatches to the appropriate method based on entry$type and returns a standardised result list.

Usage

compute_table(entry, design, data = NULL)

Arguments

Value

A list with:

id

Table identifier.

title

Table title.

type

Table type.

available

Logical: TRUE if the required variable(s) exist.

results

A list of data frames: For proportion: total, by_sex, by_age (each with estimate, lower, upper). For mean: total, by_sex, by_age (each with estimate, lower, upper). For category: total, by_sex, by_age (each with level, estimate, lower, upper). For cascade: named list of proportion results.

Description

Calculates prevalence of tobacco use from a survey design object. Computes proportions of current and daily tobacco use, stratified by sex and age group where available.

Usage

compute_tobacco_indicators(design)

Arguments

Details

When both smoking and smokeless tobacco variables are present, current_tobacco_any (either smoking or smokeless) is preferred as the headline tobacco indicator. The function also reports current_smoker and current_smokeless separately if available.

Value

A named list of survey estimates. Each element contains proportion estimates (as tibble with columns: estimate, lower, upper, etc.) for:

• current_tobacco_any_total/by_sex/by_age: any current tobacco use (smoking or smokeless) – preferred headline variable

• current_tobacco_total/by_sex/by_age: current tobacco smoking

• current_smoker_total/by_sex/by_age: current smoker

• current_smokeless_total/by_sex/by_age: current smokeless tobacco

• daily_tobacco_total/by_sex/by_age: daily tobacco use (only elements for variables present in design are returned)

See Also

compute_all_indicators()

Examples

test_data <- generate_test_data(n = 500, seed = 42)
  cols <- detect_steps_columns(test_data)
  clean <- clean_steps_data(test_data, cols)
  design <- setup_survey_design(clean)
  tobacco_results <- compute_tobacco_indicators(design)

Description

Tries to find a column in the data matching one of several candidate names (case-insensitive).

Usage

detect_col(data, candidates, label = NULL)

Arguments

Value

The matched column name (character) or NULL.

Description

Scans a data frame for standard WHO STEPS variable names across versions 3.1 and 3.2. Aliases are listed in priority order: the first match wins, so put the most specific / unambiguous name first.

Usage

detect_steps_columns(data)

Arguments

Details

WHO STEPS reorganised variable codes between v3.1 and v3.2:

v3.1 / Epi Info codes (still common in many country datasets): B1-B6 = blood-pressure readings, B7 = BP meds, C1 = fasting glucose, C5 = DM meds, C6 = total cholesterol, C10 = chol meds, M1 = height, M2 = weight, M3 = waist.

v3.2 instrument codes: M4a/M5a/M6a = SBP readings, M4b/M5b/M6b = DBP readings, M7 = BP meds, M11 = height, M12 = weight, M14 = waist, M15 = hip, B5 = fasting glucose, B6 = DM meds, B8 = total cholesterol, B9 = chol meds, B16 = triglycerides, B17 = HDL cholesterol, C1 = sex, C3 = age.

The function includes aliases for both versions so datasets from either instrument version are detected automatically.

Value

A named list of detected column names (or NULL for missing).

Description

Creates a realistic simulated dataset matching WHO STEPS survey structure. Includes sampling design variables, demographics, and measures from all three steps (behavioural, physical, biochemical).

Usage

generate_test_data(n = 3000, seed = 42)

Arguments

Details

Simulation parameters are realistic for low-middle income settings:

• Tobacco prevalence: 32% males, 8% females

• Alcohol current use: 55% males, 28% females

• Heavy episodic drinking: 35% of drinkers

• Physical activity: MET-minutes/week, mean 1800, SD 1200

• Diet: Fruit/veg days and servings per day (0-7, 1-5)

• BP increases with age; medication prevalence 12%

• Glucose: mean 5.2 mmol/L, increases with age

• Total cholesterol: mean 4.8 mmol/L

Use this function for:

• Testing the STEPS pipeline

• Developing reports before real data arrives

• Training analysts on the analysis system

Value

A data frame with n rows and the following columns:

• stratum: Strata identifier (S1-S5)

• psu: Primary sampling unit (PSU1-PSU40)

• wt_final: Final analysis weight

• sex: Sex (1=Male, 2=Female)

• age: Age in years (18-69)

• Step 1 (behavioural): t1, t2 (tobacco), a1, a5 (alcohol), met_total (physical activity), d1-d4 (diet)

• Step 2 (physical): m1 (height), m2 (weight), m3 (waist), b1-b6 (blood pressure), b7 (BP medication)

• Step 3 (biochemical): c1_mmol (glucose), c5 (DM meds), c6 (cholesterol), c10 (cholesterol meds)

Examples

# Generate smaller dataset for quick testing
  test_data <- generate_test_data(n = 500, seed = 123)
  head(test_data)

Description

Get table registry entries by section

Usage

get_registry_by_section(section = NULL)

Arguments

Value

A filtered list of registry entries.

Description

Get table registry entries by step

Usage

get_registry_by_step(step)

Arguments

Value

A filtered list of registry entries.

Description

Reads a raw STEPS data file (CSV, Excel, Stata, or SPSS) and standardises column names to lowercase with underscores.

Usage

import_steps_data(path)

Arguments

Value

A data frame with cleaned column names.

Examples

## Not run: 
raw <- import_steps_data("data/raw/steps_data.csv")

## End(Not run)

Description

List all available sections in the registry

Usage

list_registry_sections()

Value

Character vector of unique section names.

Description

Creates a tile heatmap showing missingness percentage by variable, grouped by STEPS domain.

Usage

plot_completeness(dq)

Arguments

Value

A ggplot object.

Description

Creates a bar chart of terminal-digit frequencies with the expected uniform line at 10 %.

Usage

plot_digit_preference(dq, measure)

Arguments

Value

A ggplot object.

Description

Creates a histogram of sampling weights with summary statistics.

Usage

plot_weights(dq, step = "weight_step1")

Arguments

Value

A ggplot object.

Description

Reads a filled-in column mapping template (Excel or CSV) and returns a named list suitable for passing to clean_steps_data(). The mapping file should have at least two columns: one with the standard variable name (column A) and one with the user's column name (column C in the template, or the third column).

Usage

read_column_mapping(path, data = NULL)

Arguments

Details

This function is the manual alternative to detect_steps_columns(). Use it when your dataset has non-standard variable names that auto-detection cannot resolve.

A blank template can be obtained from system.file("templates", "column_mapping_template.xlsx", package = "stepssurvey") or downloaded from the Shiny app.

The function ignores domain-header rows (rows where column A is all-caps with no entry in column C) and skips any row where the user's column name is blank.

Value

A named list where names are standard variable identifiers (e.g. "age", "sbp1") and values are the corresponding column names in the user's dataset. Unmapped variables are set to NULL.

Examples

## Not run: 
  cols <- read_column_mapping("my_mapping.xlsx")
  raw  <- import_steps_data("survey.dta")
  clean <- clean_steps_data(raw, cols)

## End(Not run)

Description

Generates a comprehensive Word document with executive summary, indicator-by-indicator analysis, and recommendations for public health action.

Usage

render_country_report(config, output_dir = tempdir())

Arguments

Details

Sections include:

• Executive summary with key findings

• Tobacco use

• Physical activity

• Overweight and obesity

• Blood pressure

• Blood glucose and cholesterol

• Recommendations for public health action

• Methodology

Requires pre-computed indicators, tables, and plots in data/processed/.

Value

Path to generated Word document (invisibly). Prints message with output location.

Description

Generates a Word document with detailed age-stratified prevalence tables for all available indicators, organized by STEPS step.

Usage

render_data_book(config, output_dir = tempdir())

Arguments

Details

Sections correspond to STEPS steps:

• Step 1: Behavioural Risk Factors (tobacco, alcohol, diet, physical activity)

• Step 2: Physical Measurements (overweight/obesity, blood pressure)

• Step 3: Biochemical (glucose, cholesterol)

Requires pre-computed tables and plots in data/processed/.

Value

Path to generated Word document (invisibly). Prints message with output location.

Description

Generates a Word document with an overview of key NCD risk factor prevalence, including summary table and sex-stratified charts.

Usage

render_fact_sheet(config, output_dir = tempdir(), format = c("html", "word"))

Arguments

Details

The fact sheet template uses pre-computed indicators, key_indicators, and plots (via .rds files in data/processed/). Requires rmarkdown, flextable, ggplot2, glue, patchwork packages.

Value

Path to generated output file (invisibly). Prints message with output location.

Description

Starts the interactive STEPS survey analysis app in the user's browser. The app provides a guided workflow: upload data, clean, set survey design, compute indicators, visualise results, and generate Word reports.

Usage

run_app(...)

Arguments

Value

A Shiny app object (invisibly). Called for its side effect of launching the application.

Examples

## Not run: 
  run_app()

## End(Not run)

Description

Imports raw data, cleans it, sets up the survey design, computes all indicators, generates publication-ready tables and plots, and optionally renders Word reports.

Usage

run_steps_pipeline(
  data_path,
  country_name = "Country Name",
  survey_year = 2024,
  age_min = 18,
  age_max = 69,
  output_dir = tempdir(),
  render_reports = TRUE,
  mapping_file = NULL
)

Arguments

Details

This is the main entry point for end-to-end STEPS analysis.

Value

A list with elements:

raw_data

Original imported data frame

clean_data

Cleaned and recoded data

cols

Detected column mapping from detect_steps_columns()

design

survey::svydesign object

indicators

List of all computed indicator results by domain

key_indicators

Summary tibble of headline estimates

tables

List of flextable::flextable objects

plots

List of ggplot2::ggplot objects

config

Configuration list from steps_config()

Examples

## Not run: 
# Auto-detect columns
result <- run_steps_pipeline("data/raw/steps_data.csv",
                             country_name = "Senegal",
                             survey_year = 2023)
result$key_indicators
result$plots$overview

# Use a custom column mapping
result <- run_steps_pipeline("data/raw/steps_data.csv",
                             country_name = "Senegal",
                             survey_year = 2023,
                             mapping_file = "my_mapping.xlsx")

## End(Not run)

Description

Exports all plots in a list to PNG files in the specified directory.

Usage

save_steps_plots(plots, output_dir = tempdir())

Arguments

Details

Files are named:

• ⁠01_overview_indicators.png⁠ (12x8 in)

• ⁠02_by_sex_dashboard.png⁠ (12x8 in)

• ⁠03_bp_by_age.png⁠ (10x6 in)

• ⁠04_obesity_by_age.png⁠ (10x6 in)

All saved at 150 dpi with white background.

Value

NULL (invisibly). Prints messages about saved files.

Description

Creates up to three survey design objects — one per WHO STEPS Step — each using the appropriate step-specific weight column (wt_step1, wt_step2, wt_step3).

Usage

setup_survey_design(data)

Arguments

Details

The returned object is a list of class "steps_designs" with elements ⁠$step1⁠, ⁠$step2⁠, ⁠$step3⁠. For backward compatibility it can also be used directly as a single design (it delegates to ⁠$step1⁠).

The function handles five design cases per step:

1. Full complex design: weights + strata + clusters

2. Weights + clusters, no strata

3. Weights + strata, no clusters

4. Weights only

5. Unweighted (simple random sampling)

Weights are used as-is without trimming, consistent with the WHO official STEPS analysis scripts.

Value

A list of class "steps_designs" with three survey::svydesign objects (step1, step2, step3).

Description

A named list of colours used in WHO STEPS reports and visualisations.

Usage

steps_colors()

Value

A named list of hex colour codes.

Examples

steps_colors()$blue

Description

Builds a configuration list that specifies data paths, design variables, and report parameters for the STEPS pipeline.

Usage

steps_config(
  data_path,
  country_name = "Country Name",
  survey_year = 2024,
  age_min = 18,
  age_max = 69,
  weight_var = "wt_final",
  strata_var = "stratum",
  cluster_var = "psu",
  bp_sbp_threshold = 140,
  bp_dbp_threshold = 90,
  bmi_overweight = 25,
  bmi_obese = 30,
  glucose_threshold = 7,
  glucose_impaired_threshold = 6.1,
  chol_threshold = 5
)

Arguments

Value

A list with elements:

• data_path: Input file path

• country_name: Country name

• survey_year: Survey year

• age_min, age_max: Age range

• weight_var, strata_var, cluster_var: Design variable names

• Threshold parameters for BP, BMI, glucose, cholesterol

Examples

## Not run: 
  cfg <- steps_config("data/steps_2023.csv", "Senegal", 2023)
  cfg <- steps_config("data/steps.csv", "Mongolia", 2019,
                      bp_sbp_threshold = 130, bp_dbp_threshold = 80)

## End(Not run)

Description

Produces a comprehensive data quality report covering digit preference, completeness, plausibility, and sampling weight diagnostics.

Usage

steps_data_quality(raw, cleaned, cols)

Arguments

Details

Digit preference / heaping is assessed using the Whipple-style heaping index: the ratio of observed frequency at a digit (0 or 5) to the expected frequency under uniform distribution. An index of 1.0 = no preference; >1.5 = moderate heaping; >2.0 = severe.

Completeness reports missing values for key STEPS variables grouped by Step (behavioural, physical, biochemical).

Plausibility counts values outside WHO-recommended ranges (e.g. height 100–250 cm, weight 20–300 kg, SBP 60–300 mmHg).

Weight diagnostics summarise the distribution of sampling weights and flag potential issues (high CV, zero/NA weights).

Value

A list of class "steps_quality" with elements:

digit_preference

Terminal-digit tables and heaping indices for physical measurements (SBP, DBP, height, weight, waist).

completeness

Per-variable missingness counts and percentages, grouped by STEPS domain.

plausibility

Summary of values outside plausible ranges.

weights

Sampling weight distribution statistics.

Description

Defines all standard tables from the WHO STEPS Epi Info report template. Each entry specifies the table metadata; generic compute and formatting functions use this registry to produce the full data book automatically.

Usage

steps_table_registry()

Value

A list of table specification lists.

Table types

proportion

Single binary indicator: % (95% CI) by age × sex. Most common type. Example: "Current smokers among all respondents."

mean

Continuous variable: mean (95% CI) by age × sex. Example: "Mean BMI (kg/m²)."

category

Multi-level factor: % per level (95% CI) by age × sex. Example: "BMI classifications (Underweight / Normal / Overweight / Obese)."

cascade

Diagnosis → treatment → control chain: multiple proportions with nested denominators. Example: "Raised BP diagnosis, treatment and control."

combined

Summary of combined risk factors: 0, 1-2, 3-5 risk factors.

Registry fields

id

Unique short identifier (e.g., "T_smoking_current").

section

Data book section (e.g., "Tobacco Use", "Blood Pressure").

step

STEPS step number (1, 2, or 3).

title

Table title as shown in the data book.

description

One-line description from the WHO template.

type

One of: "proportion", "mean", "category", "cascade", "combined".

variable

Column name(s) in the cleaned data frame to analyse. For proportion: single logical variable. For mean: single numeric variable. For category: single factor variable. For cascade: named list of logical variables.

denominator

NULL (= all respondents) or column name for subsetting (e.g., "current_alcohol" to restrict to drinkers).

levels

For category type: named character vector of level labels.

epi_info

Epi Info program name(s) for reference.

unit

Display unit (e.g., "%", "mmHg", "cm", "kg/m²", "mmol/L").

questions

STEPS instrument question codes used.

sex_panels

Logical. TRUE = 3 panels (Men/Women/Both); FALSE = 2 panels (Men/Women only, e.g., height/weight means). Default TRUE.

Description

Calculates weighted means with 95% confidence intervals for a continuous variable, optionally stratified by a grouping variable.

Usage

svymn(formula, design, by = NULL, na.rm = TRUE)

Arguments

Value

A data frame with columns:

• estimate: estimated mean

• lower: 95% CI lower bound

• upper: 95% CI upper bound

• se: standard error

• If by is specified: grouping column(s) prepended

Description

Calculates weighted proportions (as percentages) with 95% confidence intervals for a yes/no variable, optionally stratified by a grouping variable.

Usage

svyprop(formula, design, by = NULL, na.rm = TRUE)

Arguments

Value

A data frame with columns:

• estimate: estimated proportion (%)

• lower: 95% CI lower bound (%)

• upper: 95% CI upper bound (%)

• se: standard error (%)

• If by is specified: grouping column(s) prepended

Description

Takes computed results from compute_table() or compute_all_tables() and produces formatted flextable objects in the standard WHO STEPS 3-panel format (Men / Women / Both Sexes).

Description

A clean, minimal ggplot2 theme styled with WHO STEPS colours.

Usage

theme_steps(base_size = 11)

Arguments

Value

A ggplot2::theme object.

Examples

library(ggplot2)
ggplot(mtcars, aes(wt, mpg)) + geom_point() + theme_steps()