mdufa: Structured Data on FDA’s Medical Device User Fee Program Performance

This package includes structured data from the FDA’s quarterly Medical Device User Fee Amendments (MDUFA) programs, which are published as PDFs on fda.gov. The package includes functions to scrape those PDFs to extract the data and includes the most recent dataset as of publication.

Aside from some spot-checks, the data provided by the mdufa R package has had limited verification and may be inaccurate. Use this information at your own risk and verify information using the reports provided directly by FDA. To facilitate this, each data point provided includes information about its source, including a link to the FDA report from which it came, the relevant page number, and more.

“Text” metrics are typically descriptions of the MDUFA performance goals themselves (e.g. “90% Within 320 FDA Days”). Because of the way these values break across lines within table cells in the PDF reports, they are more difficult to retrieve and may be more likely to be incomplete or to be inaccurate. Similarly, some tables in the reports have footnotes, and they are not included in this dataset.

If you find a problem in this dataset, please report it here.

Please carefully read the license for additional important information.

Installation

Install the development version from GitHub with:

# install.packages("devtools")
devtools::install_github("bjoleary/mdufa")

Examples

You can access data from MDUFA 3 using mdufa::mdufa3:

library(mdufa)
library(dplyr)
#> 
#> Attaching package: 'dplyr'
#> The following objects are masked from 'package:stats':
#> 
#>     filter, lag
#> The following objects are masked from 'package:base':
#> 
#>     intersect, setdiff, setequal, union
set.seed(1)
mdufa::mdufa3 %>% 
  select(organization, program, metric_type, performance_metric, fy, value) %>% 
  sample_n(5) %>% 
  print()
#> # A tibble: 5 × 6
#>   organization program metric_type performance_metric                fy    value
#>   <chr>        <chr>   <chr>       <chr>                             <chr> <chr>
#> 1 OIR          PMA     integer     Number of PMAs filed              2014  0    
#> 2 DSD          510(k)  double      Mean FDA days for submissions th… 2017  0    
#> 3 ODE          510(k)  integer     Number of Withdrawals             2014  175  
#> 4 DCTD         510(k)  integer     80th Percentile Total days to MD… 2013  245  
#> 5 CDRH         PMA     integer     Panel-Tracked Supplements (Panel… 2015  0

Data from MDUFA 5 is available using mdufa::mdufa5. MDUFA 4 is also available using mdufa::mdufa4 and data from MDUFA 2 is available using mdufa::mdufa2.

Metrics can be assessed and graphed over time:

library(ggplot2)
data <- 
  mdufa::mdufa4 %>% 
  mdufa::filter_metrics("double") %>%
  filter(
    report_date == max(report_date),
    organization == "CDRH",
    program == "510(k)",
    performance_metric == "Average Number of Total Days to MDUFA IV Decision"
  ) %>% 
  dplyr::mutate(
    fy = as.integer(fy)
  )
graph_title <- 
  paste(data$performance_metric %>% unique(), "by Fiscal Year Received")
x_label <- "Fiscal Year Received"
y_label <- data$performance_metric %>% unique()
graph_caption <- 
  paste0(
    "Source: ",
    data$report_description %>% unique(),
    "\nAvailable at: ",
    data$report_link %>% unique()
  )
graph <- 
  ggplot(
    data = data,
    mapping = 
      aes(
        x = fy,
        y = value
      )
  ) +
  geom_point() +
  theme_classic() +
  scale_y_continuous(limits = c(0, NA)) + 
  labs(
    title = graph_title,
    y = y_label,
    x = x_label,
    caption = graph_caption
  )
plot(graph)

Exporting the data

In many cases, you may wish to locally save the data from a MDUFA report into a commonly used format, such as excel. This can be accomplished using the data already included in the package:

mdufa::export_excel(
  data = mdufa::mdufa4,
  filepath = "mdufa4_quarterly_performance.xlsx"
)

Building the datasets for yourself

Code to build the datasets for yourself can be found in data-raw/.