Heritage Health Prize


IMPORTANT NOTE: The information provided below is intended only to provide general guidance to participants in the Heritage Health Prize Competition and is subject to the Competition Official Rules. Any capitalized term not defined below is defined in the Competition Official Rules. Please consult the Competition Official Rules for complete details.

Updated as of May 4, 2011


Why the Competition?

Dr. Richard Merkin, President and CEO of Heritage Provider Network, Inc. ("HPN") developed the U.S. $3 million Heritage Health Prize Competition (the "Competition") to promote innovation through predictive modeling, challenging the brightest minds in the world to build an algorithm to find a solution to the approximately U.S. $2 trillion dollars we're spending on health care in the U.S. We believe it's currently the world's largest health prize for predictive modeling, larger than The Nobel Prize for Medicine or The Gates Prize for International Health.

It is truly a visionary approach to saving the estimated US $30 billion spent by Americans in avoidable hospitalizations, changing the approach to healthcare in a global manner. Doing so would improve the health care of all Americans, while at the same time lowering the cost of care for the average American family. If the Competition is successful, HPN expects it will be able to use the winning algorithm to both keep people healthy and out of the hospital and lower the cost of care.

What is the Heritage Provider Network?

Heritage Provider Network is one of the largest integrated physicians networks in the world, providing health care for approximately 700,000 Californians. HPN connects patients with physicians, hospitals and specialists to provide quality care in a cost effective manner. HPN is a group of physicians, not an insurance company. HPN does not sell insurance.

How long will the Competition last?

The Competition begins on April 4, 2011 and ends at 6:59:59 UTC on April 4, 2013.

How do I enter the Competition?

To register, visit http://www.heritagehealthprize.com (the "Website") or www.kaggle.com and follow the onscreen instructions to complete and submit your registration. You may register at any time prior to 6:59:59 UTC on October 4, 2012.

What am I being asked to predict?

You're being asked to predict the number of days HPN members will spend in a hospital in a given year (Y4 or Y5, as applicable). For the purposes of the Competition, a hospitalization includes both inpatient admissions and emergency room visits.


Can anyone enter the Competition?

You must be of at least the age of majority or older in your place of residence as of the time that you register and you cannot be a resident of Cuba, Sudan, Iran, North Korea, Syria or any other country designated by the United States Treasury's Office of Foreign Assets Control. See http://www.treasury.gov/resource-center/sanctions/Programs/Pages/Programs.aspx for a list of sanction programs by country.

Also, if you are an officer, director, employee or advisory board member of Heritage Provider Network, Inc., Kaggle Pty Ltd, Children's Hospital of Eastern Ontario Research Institute, Inc., University of Ottawa, University of Maryland, Baltimore County or Privacy Analytics Inc. or one of their respective affiliates, agents, judges and advertising and promotion agencies -- or are an immediate family member or member of the same household of someone who is -- you are not eligible to participate in the Competition.

Am I eligible to enter the Competition if I have/had a business relationship with HPN?

Yes. If you are or were a third party service provider, consultant or contractor of HPN, you may participate in the Competition; however, you may not use in the Competition any data or information received in connection with your work for or on behalf of HPN.


Do I need to be part of a Team to participate?

No. You may participate individually or as a member of a Team but not both.

Is there a limit on the number of people who may be on a Team?

Yes. A Team can have up to eight (8) members. Each Team member must register individually.

Can I be part of more than one Team?

No, you can't. We limit you to participating individually or as a member of only one Team because we have a strict one Entry per day limit. If you are part of several Teams or register with several accounts, you can circumvent this limit. If we learn that a participant is part of more than one Team, we reserve the right to disqualify him/her and remove his/her Entries. We feel strongly that every Entrant should compete on a level playing field.

How do I form a Team?

One Team member must serve as the Team leader. Initially, the Team leader will be the Team member who first completes the ‘Team Wizard’ on the Competition Website. You will be prompted to complete the Team Wizard just before you submit your first Entry. You will not be considered a member of a Team unless and until you confirm your Team membership by responding to the Team notification message available through your Account.

Can the Team leader change?

Yes. To change the Team leader, please use the Website's Contact Us Form.

What are Team Mergers and why are you limiting them?

Team mergers allow two separate Teams to combine to participate in the Competition as a single Team. They are a great feature of the Competition because they permit participants to collaborate with others with whom they wouldn't otherwise work and to learn new techniques.

However, we must balance the benefits of Team mergers with the need to keep the Competition as fair as possible. We want everybody to compete on a level playing field so Team mergers will be considered on a case-by-case basis only when requested. They generally will be permitted when we determine that the merger will not result in the merged Teams getting more Entries to win one of the Competition's seven cash prizes than other participants.


When will I get the Competition Data Sets?

You will get the Data Sets after completing the registration process. Data Sets will be released to Entrants on the Website as follows:

April 4, 2011 Claims Table - Y1 and DaysInHospital Table - Y2
May 4, 2011 All other Data Sets except Labs Table and Rx Table
June 4, 2011 Labs Table and Rx Table

Why has certain identifying information been removed from the Data Sets?

The identity of individual patients and health care providers, as well as other individual identifiable information, has been removed from the Data Sets to protect the privacy of those involved and to comply with applicable law.

How should I use the Data Sets?

The Data Sets should be used only to participate in the Competition, including but not limited to developing your Competition Entries and participating in the forum discussions on the Website.

How should I maintain the Data Sets?

You must maintain customary and prudent medical data security and privacy practices (including but not limited to data confidentiality, storage and encryption practices) with respect to the Data Sets and other data used in connection with the Competition. If we determine that you aren't, you may be disqualified.

Can I use data other than the Data Sets?

Yes. You may use data other than the Data Sets to develop and test your/your Team’s Prediction Algorithm and Entries until 11:59:59 UTC on April 4, 2012 if (a) the data are freely available to all other Entrants and (b) you publish the data or a link to the data in the External Data portion of the Forum section of the Website within one (1) week of your submission of an Entry using the data.

You may not, however, link the Data Sets to records in other external databases such that new demographic, socioeconomic or clinical information about the members in the Data Sets is gained. HPN reserves the right in its sole discretion to disqualify any Entrant who it discovers has undertaken or attempted to undertake such linking of the Data Sets.

You may not use any data other than the Data Sets after 11:59:59 UTC on April 4, 2012 without HPN's prior approval.

Can I use the Data Sets for research purposes?

HPN does not permit use of the Data Sets for any purpose other than participation in the Competition without HPN's prior written approval.

Why do the DaysInHospital_Yx Tables contain only a subset of all members?

The DaysInHospital_Yx Tables only contain the members who are eligible to make a claim in Yx and the 12-month period prior to Yx. For these purposes, "eligible to make a claim" refers to individuals who were HPN members in both those years. For example, the DaysInHospital_Y4 Table contains HPN members who each were eligible to make a claim in Y4 and Y3. A member can become ineligible to make a claim if he/she ceases to be a HPN member.


What is an "Entry"?

"Entry" means the data submitted via the Website in the manner and format specified on the Website on an Entry form.

How do I submit Entries?

Each Entry must be uploaded to the Website in the manner and format specified on the Website.

How many Entries can I submit?

You or your Team may submit one Entry each day during the Competition period, beginning on May 4, 2011.

Are there any fees for submitting Entries?

No. You can participate in the Competition without paying any fee.

Will all my Entries be considered for all of the prizes?

No. You must designate which Entries you want considered for each prize. You may (but are not required to) designate one Entry for consideration for each Milestone Prize round and up to five Entries for consideration for the Grand Prize. If you don't designate five Entries for consideration for the Grand Prize, your five Entries with the lowest Prediction Error Rates (as described in Rule 12 of the Competition Official Rules) on the Leaderboard will be automatically designated for Grand Prize consideration.

Can I change my Entry designations?

Yes. Entry designations for consideration for a prize can be changed any time prior to the applicable designation deadline.

What are the deadlines for designating Entries for prizes?

All designations for Milestone Prizes Entries must be made by 06:59:59 (UTC) on the following dates:

Round 1 Milestone Prizes 06:59:59 UTC on August 31, 2011
Round 2 Milestone Prizes 06:59:59 UTC on February 13, 2012
Round 3 Milestone Prizes 06:59:59 UTC on September 4, 2012

All Grand Prize Entry designations must be made by 06:59:59 UTC on April 4, 2013.

What is a "Prediction Algorithm"?

A "Prediction Algorithm" is the algorithm used to produce the data in an Entry taken as a whole (i.e., its particular total configuration) but does not include individual components of the Prediction Algorithm or tools used for analysis or development of the Prediction Algorithm.

Must my Prediction Algorithm and Entries be my (or my Team’s) original work?

Yes. Your Prediction Algorithm and Entries must be the exclusive original work of you (or if applicable, your Team), must not have been previously published or won any other prize and must not violate any laws or third party rights.

Should I use a version control system in developing my Prediction Algorithm?

We recommend that you use some kind of version control system (like git) to help ensure that, if you are a conditional prize winner, the judges will be able to use your Prediction Algorithm to reproduce your Entries.

What is the Prediction Error Threshold?

The Prediction Error Threshold is the maximum Prediction Error Rate required to win the Grand Prize. The Prediction Error Threshold will be published on the Website on May 4, 2011. The Grand Prize will be awarded to the Entrant that develops the Prediction Algorithm that produces the eligible Grand Prize Entry with the lowest Prediction Error Rate that is at least as low as the Prediction Error Threshold as of 06:59:59 UTC on April 4, 2013.

What if no Grand Prize Entry has a Prediction Error Rate lower than the Prediction Error Threshold?

If no Grand Prize Entry has a Prediction Error Rate that is lower than the Prediction Error Threshold as of 6:59:59 UTC on April 4, 2013, the US $3 million Grand Prize will not be awarded and the US $500,000 Consolation Prize will be awarded to the eligible Grand Prize Entry having the lowest Prediction Error Rate.

If no Grand Prize Entry meets the Accuracy Threshold as of 6:59:59 UTC on April 4, 2013, the U.S. $3 million Grand Prize will not be awarded and the US $500,000 Consolation Prize will be awarded to the eligible Grand Prize Entry having the most accurate prediction score.

Do Milestone Prize Entries need to meet the Prediction Error Threshold?


If my Entry is selected as a conditional winner, what else will I need to provide to verify my Entry?

If your Entry is selected as a conditional winner, you must provide the code for the Prediction Algorithm that produced your Entry and any requested back-up documentation. You also may be asked to submit computer hardware or virtual machine instance that runs the code.

What can I/we use to develop my/our Prediction Algorithm?

You may use any algorithm and/or analytical and developmental tool, including open source tools, pre-existing algorithms and other items of your choosing.

What should I do if a component of my/our Prediction Algorithm belongs to someone else?

If any component of your Prediction Algorithm is legally available only through a fee-based license and/or for non-commercial use, then you must contact us for approval before including the component in your Prediction Algorithm. You do not need to obtain our permission for use of any component available pursuant to an OSI-approved license listed at http://www.opensource.org/licenses/alphabetical.

If I’m not (or my Team isn’t) a conditional prize winner, might I still need to submit the Prediction Algorithm to HPN?

We may request that you submit your (or your Team’s) Prediction Algorithm to us at any time. This will help us to preserve the integrity of the Leaderboard and otherwise prevent fraud so that the Competition is as fair as possible for all Entrants.

Will HPN have the right to use my Entries and Prediction Algorithm?

Yes. By registering in the Competition, you grant HPN an exclusive, worldwide, perpetual, royalty free license to use and grant others the right to use your Entries (see description above) and the Prediction Algorithm (see description above) used to produce your Entries. Such use may be for any commercial or other purpose and won’t require your further approval or any payment to you.

Will I have the right to use my Entries and Prediction Algorithm?

Yes. You may use your Entries and the Prediction Algorithm used to produce your Entries but such use must be solely for your own patient management and internal business purposes. You may not transfer or grant any rights to your Entries or Prediction Algorithm to any other entity or person.

Why is HPN seeking to control my Prediction Algorithm?

HPN wants to help ensure that Prediction Algorithms are used to promote better patient health care and not for improper purposes.


Who is judging the Competition?

Judging will be conducted by a panel of independent, qualified experts.

When will the Leaderboard be available?

From May 4, 2011 through the end of the Competition.

Do Entrants winning on the Leaderboard receive any prize?

The Leaderboard is intended only to give you an indication of how you are performing relative to other participants who have submitted Entries. Standings on the Leaderboard don't determine prize winners.

What if there are ties?

If any two or more Prediction Algorithms produce eligible Entries with the same lowest prediction score, the Entry received first will be deemed the winner of the applicable prize.


What can I win?

The Competition offers a Grand Prize of US $3,000,000 (or if no Grand Prize Entry produces a Prediction Error Rate lower than the Prediction Error Threshold, a Consolation Prize of US $500,000) and six Milestone Prizes:

Round 1 Milestone Prizes First Prize: U$30,000
Second Prize: US $20,000
Round 2 Milestone Prizes First Prize: US $50,000
Second Prize: US $30,000
Round 3 Milestone Prizes First Prize: US $60,000
Second Prize: US $40,000

What if my Team wins?

If a Team wins a prize, all Team members must submit a single written statement describing how the prize is to be allocated among the Team members. If the Team fails to submit the statement within 30 days after HPN requests it, HPN will distribute the prize to Team members in equal shares.


Who is responsible for Forum posts?

You are responsible for your own Forum posts, although HPN reserves the right to remove posts which contain inappropriate, offensive or illegal material.