Overview

The KPHD package is a comprehensive toolkit designed by and for epidemiologists at Kitsap Public Health District. It provides visualization tools following KPHD style guidelines alongside practical epidemiological functions to streamline public health data analysis, reporting, and surveillance workflows.

Key Features

• KPHD Style Guide: Brand-compliant themes, color palettes, and typography
• Epidemiological Tools: Pending
• Data Privacy: Primary and secondary suppression for protecting small cell counts
• Surveillance Workflows: Pending
• Spatial Analysis: Pending
• Reproducible Reports: Pending

Installation

Development Version

Install the development version from GitHub:

# install.packages("devtools")
devtools::install_github("KPHD-Assessment-Epidemiology/KPHD")

Required Dependencies

The package requires the following packages: - ggplot2 - for plotting - showtext and sysfonts - for Poppins font support - grDevices - for color interpolation

Quick Start

library(KPHD)
library(ggplot2)

# Setup KPHD theme
setup_kphd_fonts()
set_kphd_theme()

# Visualization example
ggplot(mtcars, aes(x = wt, y = mpg, color = factor(cyl))) +
  geom_point(size = 3) +
  scale_color_kphd() +
  labs(
    title = "Fuel Economy by Weight",
    subtitle = "Using KPHD theme and colors",
    x = "Weight (1000 lbs)",
    y = "Miles per Gallon",
    color = "Cylinders"
  )

Color Palettes

View All Palettes

show_kphd_palettes()

Available Palettes

• primary: Main KPHD brand colors (colorblind-friendly)
• sequential: Marine blue gradients for continuous data
• diverging: Purple to marine blue through neutral
• categorical: Extended palette for multiple categories
• colorblind_safe: Colorblind-accessible colors
• high_contrast: Maximum accessibility palette

Using Color Palettes

# Get specific palette
colors <- get_kphd_colors("primary")

# Get marine blue (primary brand color)
marine_blue <- get_kphd_colors("marine_blue")

# Get interpolated colors
colors_10 <- get_kphd_colors("sequential", n = 10)

# Use in ggplot2
ggplot(data, aes(x = x, y = y, fill = category)) +
  geom_col() +
  scale_fill_kphd(palette = "primary")

# Continuous scales
ggplot(data, aes(x = x, y = y, fill = value)) +
  geom_tile() +
  scale_fill_kphd_c(palette = "sequential")

Theme Customization

Default KPHD Theme

# Set as default for all plots in session
setup_kphd_fonts()
set_kphd_theme()

# All subsequent plots use KPHD theme
ggplot(data, aes(x, y)) + geom_point()

Per-Plot Customization

# Use without setting as default
ggplot(data, aes(x, y)) +
  geom_point() +
  theme_kphd(base_size = 14)

# Without Poppins font
ggplot(data, aes(x, y)) +
  geom_point() +
  theme_kphd(base_family = "sans")

Data Suppression for Privacy

The kphd_suppress() function implements primary and secondary suppression to protect privacy in public health data.

Basic Usage

# Example data with small counts
health_data <- data.frame(
  region = c("North", "South", "East", "West", "Central"),
  year = rep(2023, 5),
  cases = c(5, 150, 8, 200, 12)
)

# Apply suppression (threshold = 10 by default)
suppressed_data <- kphd_suppress(
  health_data,
  count_var = "cases",
  group_vars = "year",
  threshold = 10
)

# View results
print(suppressed_data)
#   region year cases suppression_flag
# 1  North 2023    NA          primary
# 2  South 2023   150             none
# 3   East 2023    NA          primary
# 4   West 2023   200             none
# 5 Central 2023   NA        secondary

Suppression Summary

kphd_suppression_summary(health_data, suppressed_data, "cases")

# === KPHD Data Suppression Summary ===
# 
# Total rows: 5 
# Rows suppressed: 3 ( 60 %)
#   - Primary suppression: 2 
#   - Secondary suppression: 1

Advanced Suppression

# Multiple grouping variables
health_data2 <- data.frame(
  region = rep(c("North", "South"), each = 3),
  age_group = rep(c("0-17", "18-64", "65+"), 2),
  cases = c(8, 45, 3, 150, 7, 89)
)

suppressed <- kphd_suppress(
  health_data2,
  count_var = "cases",
  group_vars = c("region", "age_group"),
  threshold = 10,
  replace_with = NA  # or use -999, 0, etc.
)

# Without suppression flags (cleaner output)
suppressed_clean <- kphd_suppress(
  health_data2,
  count_var = "cases",
  group_vars = c("region", "age_group"),
  flag_suppressed = FALSE
)

Complete Examples

Example 1: Outbreak Investigation with Epidemic Curve

library(KPHD)
library(ggplot2)
library(dplyr)

# Setup
setup_kphd_fonts()
set_kphd_theme()

# Outbreak data
outbreak_data <- data.frame(
  onset_date = seq(as.Date("2024-01-01"), as.Date("2024-02-15"), by = "day"),
  cases = c(0, 0, 1, 2, 3, 5, 8, 12, 15, 18, 22, 25, 20, 18, 15, 
           12, 10, 8, 6, 5, 4, 3, 2, 2, 1, 1, 0, 0, 0, 0, 
           rep(0, 16))
)

# Create epidemic curve
ggplot(outbreak_data, aes(x = onset_date, y = cases)) +
  geom_col(fill = get_kphd_colors("marine_blue")) +
  scale_x_date(date_breaks = "1 week", date_labels = "%b %d") +
  labs(
    title = "Foodborne Illness Outbreak - Epidemic Curve",
    subtitle = "Cases by date of symptom onset",
    x = "Date of Onset",
    y = "Number of Cases",
    caption = "Data: KPHD Disease Investigation | Peak: January 11, 2024"
  ) +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))

# Calculate key metrics
peak_date <- outbreak_data$onset_date[which.max(outbreak_data$cases)]
total_cases <- sum(outbreak_data$cases)
attack_rate <- (total_cases / 500) * 100  # If exposed population = 500

cat("\nOutbreak Summary:\n")
cat("Total cases:", total_cases, "\n")
cat("Peak date:", format(peak_date, "%B %d, %Y"), "\n")
cat("Attack rate:", round(attack_rate, 1), "%\n")

Example 2: Disease Rate Calculations with Suppression

# Disease surveillance data by county
disease_data <- data.frame(
  county = c("Kitsap", "Jefferson", "Clallam", "Mason", "Thurston"),
  cases = c(145, 8, 67, 12, 234),
  population = c(275000, 32000, 78000, 65000, 295000),
  year = rep(2023, 5)
)

# Calculate rates per 100,000
disease_data <- disease_data %>%
  mutate(
    rate = (cases / population) * 100000,
    rate_lower = # Calculate 95% CI lower bound
      (qchisq(0.025, 2 * cases) / 2 / population) * 100000,
    rate_upper = # Calculate 95% CI upper bound
      (qchisq(0.975, 2 * (cases + 1)) / 2 / population) * 100000
  )

# Apply suppression
disease_data_suppressed <- kphd_suppress(
  disease_data,
  count_var = "cases",
  threshold = 10
)

# Visualize (non-suppressed only)
disease_data_plot <- disease_data_suppressed %>%
  filter(suppression_flag == "none")

ggplot(disease_data_plot, aes(x = reorder(county, rate), y = rate)) +
  geom_col(fill = get_kphd_colors("marine_blue")) +
  geom_errorbar(
    aes(ymin = rate_lower, ymax = rate_upper),
    width = 0.3,
    color = get_kphd_colors("additional_3")
  ) +
  coord_flip() +
  labs(
    title = "Disease Rates by County, 2023",
    subtitle = "Rate per 100,000 population with 95% confidence intervals",
    x = NULL,
    y = "Rate per 100,000",
    caption = "* Counties with <10 cases suppressed for privacy"
  )

Example 3: Age-Specific Analysis

# Age group analysis
age_data <- data.frame(
  age_group = c("0-4", "5-17", "18-24", "25-44", "45-64", "65-74", "75+"),
  cases = c(12, 45, 38, 156, 234, 189, 98),
  population = c(15000, 45000, 35000, 85000, 75000, 35000, 25000)
) %>%
  mutate(
    rate = (cases / population) * 100000,
    percent = (cases / sum(cases)) * 100
  )

# Create population pyramid-style comparison
ggplot(age_data, aes(x = age_group, y = rate)) +
  geom_col(fill = get_kphd_colors("marine_blue")) +
  geom_text(
    aes(label = sprintf("%.1f", rate)),
    hjust = -0.2,
    color = get_kphd_colors("marine_blue"),
    size = 3.5
  ) +
  coord_flip() +
  scale_y_continuous(expand = expansion(mult = c(0, 0.15))) +
  labs(
    title = "Disease Rate by Age Group",
    subtitle = "Cases per 100,000 population, 2023",
    x = "Age Group",
    y = "Rate per 100,000"
  )

Example 4: Time Trend Analysis with Moving Average

# Monthly case data over 3 years
monthly_data <- data.frame(
  month = seq(as.Date("2021-01-01"), as.Date("2023-12-01"), by = "month"),
  cases = c(
    # 2021
    45, 52, 48, 55, 62, 58, 65, 72, 68, 55, 48, 42,
    # 2022  
    38, 42, 45, 52, 58, 65, 72, 78, 85, 92, 88, 82,
    # 2023
    75, 72, 68, 65, 70, 75, 82, 88, 92, 95, 90, 85
  )
) %>%
  mutate(
    # Calculate 3-month moving average
    ma_3month = zoo::rollmean(cases, k = 3, fill = NA, align = "right")
  )

# Visualize trend
ggplot(monthly_data, aes(x = month)) +
  geom_line(
    aes(y = cases, color = "Monthly Cases"),
    linewidth = 0.8,
    alpha = 0.5
  ) +
  geom_line(
    aes(y = ma_3month, color = "3-Month Moving Average"),
    linewidth = 1.2
  ) +
  scale_color_manual(
    values = c(
      "Monthly Cases" = get_kphd_colors("secondary"),
      "3-Month Moving Average" = get_kphd_colors("marine_blue")
    )
  ) +
  scale_x_date(date_breaks = "3 months", date_labels = "%b %Y") +
  labs(
    title = "Disease Trend Analysis, 2021-2023",
    subtitle = "Monthly cases with 3-month moving average",
    x = "Month",
    y = "Number of Cases",
    color = NULL
  ) +
  theme(
    legend.position = "bottom",
    axis.text.x = element_text(angle = 45, hjust = 1)
  )

Example 5: Multi-Region Comparison with Small Count Suppression

# Regional surveillance data
regional_data <- expand.grid(
  region = c("North", "South", "East", "West"),
  year = 2019:2023
) %>%
  mutate(
    cases = c(
      # North
      45, 52, 48, 8, 55,
      # South
      78, 85, 92, 88, 95,
      # East
      5, 4, 6, 3, 7,
      # West
      65, 72, 68, 70, 75
    ),
    population = rep(c(50000, 85000, 15000, 70000), each = 5),
    rate = (cases / population) * 100000
  )

# Apply suppression by region-year
regional_suppressed <- kphd_suppress(
  regional_data,
  count_var = "cases",
  group_vars = c("region", "year"),
  threshold = 10
)

# Visualize trends (non-suppressed only)
regional_plot <- regional_suppressed %>%
  filter(suppression_flag == "none")

ggplot(regional_plot, aes(x = year, y = rate, color = region, group = region)) +
  geom_line(linewidth = 1.2) +
  geom_point(size = 3) +
  scale_color_kphd(palette = "primary") +
  labs(
    title = "Regional Disease Rate Trends",
    subtitle = "Rates per 100,000 population (suppressed cells excluded)",
    x = "Year",
    y = "Rate per 100,000",
    color = "Region",
    caption = "Note: Some data points suppressed for privacy (n<10)"
  ) +
  theme(legend.position = "bottom")

# Summary statistics
kphd_suppression_summary(regional_data, regional_suppressed, "cases")

Function Reference

Visualization & Style

• theme_kphd() - KPHD ggplot2 theme
• set_kphd_theme() - Set KPHD theme as default
• setup_kphd_fonts() - Load Poppins font from Google Fonts
• get_kphd_colors() - Get KPHD color palette
• show_kphd_palettes() - Display all available palettes
• scale_color_kphd() - Color scale for discrete variables
• scale_fill_kphd() - Fill scale for discrete variables
• scale_color_kphd_c() - Color scale for continuous variables
• scale_fill_kphd_c() - Fill scale for continuous variables

Data Privacy & Suppression

• kphd_suppress() - Apply primary and secondary suppression
• kphd_suppression_summary() - Generate suppression report

Epidemiological Calculations (Pending)

Time Series & Surveillance (Pending)

Data Cleaning & Validation (Planned / Pending)

• kphd_clean_dates() - Standardize date formats
• kphd_validate_ages() - Validate and categorize age data
• kphd_geocode() - Geocode addresses for mapping

Reporting & Tables (Planned / Pending)

• kphd_table_one() - Generate demographic tables
• kphd_crosstab() - Create publication-ready crosstabs
• kphd_summary_stats() - Calculate common summary statistics
• kphd_export_table() - Export tables to Word/Excel with formatting

KPHD Brand Colors

Color Name	Hex Code	RGB	Usage
Marine Blue	#004960	0, 74, 97	Primary brand color
Secondary	#54b4b7	85, 181, 183	Secondary accent
Background	#fbf0d7	252, 240, 216	Light backgrounds
Additional 1	#245c85	40, 99, 145	Data visualization
Additional 2	#DAFF00	218, 255, 0	High contrast
Additional 3	#56215e	87, 33, 94	Purple accent

Development Status

This package is under active development. Current status:

✅ Completed

• Core theme functionality with Poppins font
• Comprehensive color palette system
• Data suppression functions (primary & secondary)
• Basic documentation and examples
• Unit tests for existing functions
• Colorblind-safe palette options

📋 Planned - Spatial Analysis

• Geographic data cleaning and validation
• Choropleth map templates
• Spatial clustering detection (SaTScan integration)
• Distance-based calculations
• ZIP code to census tract crosswalks

📋 Planned - Data Management

• Date standardization and validation
• Age calculation and categorization
• Record deduplication algorithms
• Data quality checks and reports

📋 Planned - Reporting

• Table One generators (demographic tables)
• Publication-ready crosstabs
• Automated summary statistics
• R Markdown templates (reports, briefs, presentations)
• Export utilities (Word, Excel, PowerPoint with KPHD branding)

Development Setup

# Clone and setup
devtools::load_all()
devtools::document()
devtools::test()
devtools::check()

Getting Help

• Email: trey.evans@kitsappublichealth.org
• Bug reports: GitHub Issues
• Questions: GitHub Discussions

Citation

If you use this package in your work, please cite:

Kitsap Public Health District (2025). KPHD: Epidemiological Tools and Style 
  Guide for Public Health Analysis. R package version 0.1.0.
  https://github.com/KPHD-Assessment-Epidemiology/KPHD

BibTeX:

@Manual{,
  title = {KPHD: Epidemiological Tools and Style Guide for Public Health Analysis},
  author = {{Kitsap Public Health District}},
  year = {2025},
  note = {R package version 0.1.0},
  url = {https://github.com/KPHD-Assessment-Epidemiology/KPHD},
}

License

This package is licensed under the MIT License. See LICENSE file for details.

Acknowledgments

• Developed by epidemiologists at Kitsap Public Health District
• Color palette design based on KPHD Style Guide
• Colorblind-safe palettes informed by the work of the paletter gallery
• Built with ggplot2, dplyr, and the tidyverse
• Font support via showtext and sysfonts

---

Kitsap Public Health District