Skip to the content.

Background

The records in this subdirectory were provided by the Pennsylvania Insurance Department Open Records team in response to a PA Right to Know Law (RTKL) request submitted to the Pennsylvania Department of Insurance. This public records request was granted only in part; the request included numerous pieces of data that were unable to be provided.

Raw Data

The data in the data/claims_denials/pa/raw subdirectory consists of pdfs that describe the following claims information, for plan years 2020 and 2021 only:

Downloads

This entire site is open source, and as such you can inspect and download any subset of the data we are releasing here from github directly.

Alternatively, we provide a few quick links below:

You can also browse individual raw pdfs directly:

Notes

Disclaimer

The records in this subdirectory are public records. To our knowledge, the records have not been modified or altered in any way, and were graciously provided as-is by the PA Insurance Department. That department is not responsible for any findings, manipulations or alterations of this data. In particular, formulations of this data that we may standardize or adapt from these raw pdfs to make them more user-friendly for consumption by automated processes is in no way vetted or validated by that department, or guaranteed to be consistent with the original data; any errors resulting from such processes are our own, and not those of the PA Department of Insurance.

Parsed Data

The raw data provided in the public records comes from tables in PDF files (one for each issuer). The data is inconvenient to work with in this format, for various reasons, and the matter is made even less convenient because the tables appear to be split across pdf pages. In order to analyze this data, we’ve scraped it from the pdf files and housed it in two centralized csvs: one for issuer-level information, and one for plan-level information. These parsed data are not public records, and were not provided by any official entity: trust them at your own risk, or validate for yourself that they were parsed correctly.

Downloads

You can download the parsed data below:

In the future, we will house the script used to parse the data from these pdfs in this repo. For now, we just provide the parsed data, since the schema for the output standard that we want to support long term is still being developed.

Acknowledgements

We thank the Pennsylvania Department of Insurance, their Open Records Office, and all of the staff involved in supporting the RTLK request we submitted. We believe this information will be useful to the public and help protect consumers, and we are grateful for the consideration and help we received from the all of the staff involved.