Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2023 Jun 5;13(1):9122.
doi: 10.1038/s41598-023-36023-5.

Predicting health outcomes in dogs using insurance claims data

Christian Debes  1 , Johannes Wowra  2 , Sarosh Manzoor  2 , Audrey Ruple  3
Affiliations

Predicting health outcomes in dogs using insurance claims data

Christian Debes et al. Sci Rep. .

Abstract

In this paper we propose a machine learning-based approach to predict a multitude of insurance claim categories related to canine diseases. We introduce several machine learning approaches that are evaluated on a pet insurance dataset consisting of 785,565 dogs from the US and Canada whose insurance claims have been recorded over 17 years. 270,203 dogs with a long insurance tenure were used to train a model while the inference is applicable to all dogs in the dataset. Through this analysis we demonstrate that with this richness of data, supported by the right feature engineering, and machine learning approaches, 45 disease categories can be predicted with high accuracy.

PubMed Disclaimer

Conflict of interest statement

This work has been funded by Fetch Forward, a subsidiary of Fetch, Inc. All authors have consulted for Fetch, Inc. and received compensation for their work.

Figures

Figure 1
Figure 1
Distribution of Age at inception.
Figure 2
Figure 2
Distribution of latest age.
Figure 3
Figure 3
Proposed approach for disease prediction.
Figure 4
Figure 4
A diagram illustrating how the training samples are generated from the disease history. The colored bars represent examples of different disease claims on the age axis.
Figure 5
Figure 5
Comparison of the test AUC scores for Extreme Gradient Boosting and Logistic Regression on different disease categories.
Figure 6
Figure 6
Comparison of the test AUC scores for Extreme Gradient Boosting, Logistic Regression and Ensemble on different diseases.
Figure 7
Figure 7
Feature importance plot for diabetes prediction.
Figure 8
Figure 8
Feature importance plot for arthritis prediction.
Figure 9
Figure 9
Partial dependence plots for arthritis prediction.
See this image and copyright information in PMC

References

    1. StrategyR, pet care - global market trajectory & analytics https://www.strategyr.com/market-report-pet-care-forecasts-global-indust... Accessed 12 January 2022.
    1. Allied market research, pet insurance market by policy coverage (accident only, accident and illness, and others), by animal type (dogs, cats, others), by sales channel (agency, broker, and others): Global opportunity analysis and industry forecast, 2021-2030. https://www.alliedmarketresearch.com/pet-insurance-market Accessed 12 January 2022.
    1. American veterinary medical association, pet ownership & demographic 2018. https://ebusiness.avma.org/ProductCatalog/product.aspx?ID=1529. Accessed 12 January 2022.
    1. Statista: Number of pet owning households in europe 2010-2020. https://www.statista.com/statistics/515192/households-owning-a-pet-europe/. Accessed 12 January 2022.
    1. Williams A, Williams B, Hansen CR, Coble KH. The impact of pet health insurance on dog owners’ spending for veterinary services. Animals. 2020;10(7):1162. doi: 10.3390/ani10071162. - DOI - PMC - PubMed

Publication types