TL;DR
This study reveals that not all medical conditions are created equal when it comes to online crowdfunding success. While certain diseases like neoplasms consistently perform well, others, like mental health disorders, can actually hinder fundraising. Crucially, the socioeconomic status of a campaign's location plays a significant role, with less deprived areas seeing disproportionately higher success, primarily driven by a larger number of donations rather than larger individual contributions.
What Problem Does This Paper Address?
Online crowdfunding platforms like GoFundMe have become a significant avenue for individuals to raise funds for medical expenses, especially in the face of rising healthcare costs and insufficient insurance coverage. However, our understanding of what drives success on these platforms is incomplete. Previous research often focused on single disease categories or narrowly defined groups, failing to capture the complexity of multimorbidity. Furthermore, while socioeconomic disparities are known to affect health outcomes, their specific impact on the success of medical crowdfunding campaigns, and how this interacts with the medical conditions themselves, remained unclear. This paper aims to bridge this gap by systematically analyzing the influence of specific medical conditions and the socioeconomic context of a campaign's location on fundraising outcomes.
Methods
The researchers employed a robust methodology to analyze a large dataset of medical crowdfunding campaigns. They utilized web scraping to collect data from GoFundMe, focusing on campaigns in the US between 2010 and 2020. A key innovation was the use of a validated natural language processing (NLP) algorithm to identify the presence of 11 broad disease categories within campaign descriptions. This allowed for a more nuanced understanding of the medical conditions being funded, moving beyond single-disease analyses.
To assess the socioeconomic context, an Area Deprivation Index (ADI) was calculated for each campaign's county of origin, incorporating factors like unemployment, poverty, and internet access. Generalized linear models (GLMs) were then used to estimate the impact of disease categories and ADI quartiles on the amount of money raised. To further dissect fundraising success, separate models were built to analyze the factors influencing the number of donations and the mean donation amount.
Datasets
The primary dataset for this study was derived from GoFundMe, a popular online crowdfunding platform. The researchers collected data on 89,645 unique medical crowdfunding campaigns created between 2010 and 2020. The socioeconomic data was sourced from the US Census Bureau's American Community Survey (ACS) for county-level analysis. No new benchmarks were introduced; rather, existing data was analyzed using novel NLP and statistical modeling approaches.
Key Insights and Ideas
- Disease Matters, But Not Equally: The specific medical condition(s) mentioned in a campaign significantly influence its fundraising success. This suggests that potential donors' perceptions and willingness to contribute are not uniform across all health issues.
- The "Worthy" Narrative: The study hints at the importance of how a campaign frames its narrative. Conditions that might be perceived as preventable or linked to lifestyle choices (like diabetes) or those associated with stigma (like mental health disorders) may face an uphill battle, potentially due to donor perceptions of "worthiness."
- Location, Location, Location: Socioeconomic deprivation is a powerful predictor of fundraising success, independent of the medical condition. This highlights how broader societal inequalities can manifest even in online charitable giving.
- Quantity Over Quality (of Donations): Fundraising success is more strongly driven by the number of donations received than by the average donation amount. This implies that reaching a wider audience and generating broad engagement is more critical than securing a few large donations.
Key Results
- Neoplasms Lead the Pack: Campaigns mentioning neoplasms (cancers) were the most frequent and saw a significant increase in fundraising (24.7% more than those without).
- Beneficial Conditions: Injuries, external causes, respiratory diseases, nervous system diseases, and infections were also associated with higher fundraising amounts.
- Detrimental Conditions: Conversely, mentioning endocrine diseases (-11.1%), mental health disorders (-7.7%), and genitourinary diseases (-6.5%) was associated with lower fundraising amounts.
- Deprivation Disadvantage: Campaigns from the least-deprived counties raised significantly more (48.3%) than those from the most-deprived counties. This effect was consistent across intermediate deprivation levels, though attenuated.
- Donation Dynamics: While most disease categories influenced the number of donations, only a few (neoplasms, injuries, mental health disorders) significantly impacted the mean donation amount. Similarly, ADI primarily affected the number of donations, with a less pronounced effect on the average donation size.
Limitations and Future Work
The authors acknowledge several limitations:
- Platform Specificity: Findings are based on GoFundMe, which may not be generalizable to all crowdfunding platforms.
- Data Availability: Campaigns can be voluntarily removed, potentially introducing selection bias.
- NLP Algorithm Limitations: The algorithm could not distinguish between incidentally mentioned conditions and those central to the campaign, nor could it capture disease severity. It also had lower precision for certain disease categories.
- Exclusion of Non-English Campaigns: Technical limitations led to the exclusion of non-English campaigns, potentially impacting generalizability.
- Excluded Medical Categories: Certain medical conditions (e.g., obstetric, gynecologic, hematologic) were excluded due to data limitations during algorithm development.
- Focus on Amount Raised: The study focused on the total amount raised rather than the percentage of the goal met, which might be more relevant to individual organizers.
- Unmeasured Factors: Many factors, such as semantic features of descriptions and social capital of organizers, were not measured.
Future work could explore these excluded categories, investigate the impact of semantic features and social capital, and analyze campaigns in other languages to provide a more comprehensive understanding of medical crowdfunding success.
Summary updated