We chose to conduct a systematic review, rather than choosing another method, for the following reasons: First, a systematic review could provide a comprehensive understanding concerning the success factors of medical crowdfunding campaigns. The comprehensive understanding could be a solid basis for further studies on this topic. Second, a systematic review could limit bias in identifying and rejecting bias by using explicit methods. Third, conclusions of the systematic review may be more reliable and accurate because it summarizes previous literature systematically. Fourth, the findings of systematic review could be used to reduce the delay between scientific discoveries and implementation. The generalizability of findings could be established by comparing studied success factors of medical crowdfunding campaigns [22].

We used PubMed, PsycINFO, Web of Science, ACM Digital Library, and ScienceDirect databases. These databases were chosen because they cover most disciplines that study medical crowdfunding, namely, medicine, information, psychology, global health, computer science, and business economics. Keywords and synonyms used in this search revolved around 2 concepts—crowdfunding, and health (Textbox 1). Database searches were conducted on June 8, 2020 by 2 authors (LZ and XH) with Library and Information Science expertise. The results were exported to Zotero (version 5.0.0.0; Corporation for Digital Scholarship) and organized into folders by database. PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) was used to document the review process was used.

Papers that focused on the success of medical crowdfunding or other factors related to the success of crowdfunding (eg, narrative strategies and ethical factors of medical crowdfunding) published between 2010 and 2020 (because medical crowdfunding emerged after 2008) within conference proceedings or journals were included, and papers without research design details or results or not be written in English were excluded.

Papers were rated for quality using the McGill Mixed Methods Appraisal Tool version 2018 [23]. Its internal reliability, usability, and content validity have been verified in several studies [24,25]. Quality criteria are applied based on the study design (qualitative research, quantitative research, mixed method research) and methodology (such as randomized controlled, nonrandomized and descriptive studies). We used criteria [23] for quantitative descriptive studies: (1) Is the sampling strategy relevant to address the quantitative research question? (2) Is the sample representative of the population understudy? (3) Are measurements appropriate? (4) Is there an acceptable response rate? We also used a quality threshold (75%, ie, meeting at least 3 criteria) defined by a previous study [26]. Papers that did not meet the threshold were excluded.

Metadata were extracted and listed in a spreadsheet (Excel, Microsoft Inc): author region, publication date, data sources, research questions, data collection settings, methods, results, discussions, conclusions, and bibliographies. To interpret our findings, all authors discussed how to present our findings systematically. Since the conception of medical crowdfunding is dynamic, we discuss the conception, then summarize bibliographic and study information. The success factors are based to the contextual structure of medical crowdfunding. Since there are 3 main actors in medical crowdfunding, and they interact with each other based on specific campaigns [7], we categorized the success factors by platforms, raisers, donors, and campaigns. Because of the heterogeneity of outcome measures in studies included in our review, it was impractical to conduct a meta-analysis. Therefore, findings were qualitative.

A total of 441 papers were identified from the 5 databases. After removing 46 duplicates, titles and abstracts were screened by 2 authors separately. When a difference in opinion arose, a third author was involved to mediate the discussion to reach agreement. Using inclusion and exclusion criteria, 27 papers were chosen for full-text review. After full-text review, 19 papers were assessed for quality, 13 papers were included (Figure 1).

Despite the acknowledgment of the benefit of medical crowdfunding in strengthening health care, there is still a lack of consensus on what constitutes crowdfunding campaigns in health care. However, to study the success factors of medical crowdfunding, we need to have a basic understanding of medical crowdfunding. Based on previous studies [6-8], we define the scope of medical crowdfunding as crowdfunding for medical expenses in hospitals, new drugs in scientific research institutions, and new treatments (eg, hospitalization expenses, scientific research funds).

Of the 13 studies (Table 1), 9 studies [4,8,12,27-32] relied entirely on crowdfunding websites such as GoFundMe. There were 4 main analytical methods used: text analysis, regression analysis, semistructured interviews, and exploratory spatial analysis. Text analysis was used most, with 4 studies using only text analysis [4,29-31], and 2 studies using text analysis partially [12,33]. Of the 13 studies, 6 studies were directly related to the success of medical crowdfunding, and the remaining 7 studies were indirectly related. The number of publications appears to be steadily increasing over time, which indicates that this field is getting more and more attention from scholars and practitioners because of the particularity of medical crowdfunding [6,20]. Most studies were from high-income countries, with the largest source of articles being the United States [4,12,27,32,33], followed by Canada [30,34].

The determinants of intention to donate in donors can be mainly divided into demographic and individual characteristics (Table 4).

Donor gender [40,41], age [29], education level [29], income [29], and geographical location [35] had significant effects on donation behaviors. Younger donors were more willingness to donate [32], and people with higher education and income level were more likely to donate [42]. Donors were more willing to contribute to crowdfunding campaigns in or near their own regions [35]; therefore, geographic inequity exists, which is compounded by social, technological, and cultural issues [4].

Individual characteristics included cognitive state and social returns. Cognitive state is state invoked in the donor in reading a campaign’s description [12]. If donors feel threatened by reading the negative campaign descriptions, they hesitate to donate. When donors are aware the importance and urgency of the raiser's medical crowdfunding campaigns, they have higher willingness to donate [43]. Social return is a major intrinsic motivation for individuals or groups to donate to medical crowdfunding campaigns [4]. People with prosocial values and who participate more in charitable activities donate more willingly [44]. Donors with intrinsic motivation have high willingness to donate.

Success factors for campaigns can be divided into 2 aspects: the format and the content (Table 5).

Format-related factors include goal amount, campaign description length, third-person description, and social media sharing. When a campaign is close to its fundraising goal, it encourages donation behavior and increases the likelihood of project success [40,45]. Longer campaign descriptions and higher goal amount were significantly associated with amount raised [12]. In addition, third-person perspective in the description can convey a patient’s positive qualities in a way that would be not acceptable in the first-person perspective due to the testimonial effect [12]. It is also critical for crowdfunding campaigns to leverage social media [35]. The amount raised was strongly correlated with updates and shares in social media [12]. One additional tweet or retweet with more personal comments could enhance the probability of success of crowdfunding campaigns [35].

Content-related factors included the narrative strategy, authenticity, credibility, being easy to understand, giving back, and depth of need. Illness narratives represent the personal story and illness experience that patients are sharing verbally or in writing [28]. The narration style not only affected the efficiency of the dissemination of health information but also heavily affected on the potential donors' cognitions, attitudes, and behaviors [28]. Authenticity, which can be conveyed by pictures of raisers that depict their medical conditions, increases the possibility of donations. Funding is more accessible when the presented narratives of campaigns were credible [46]. To demonstrate the credibility of campaigns, many methods, including collective endorsements, presenting details of external financial support, displaying off-site verification details (of ailment, incident, and treatment), providing verification of fundraiser and beneficiary identities, and using a popular and trusted platform, can be used [33].

Easy-to-understand information in the campaign description gives potential donors a clear picture of the campaign [31]. In addition, portraying the beneficiary as someone who selflessly gives back to society not only helps establish the positive image of raisers, but also, inspires the donors themselves [30]. Patients in urgent need of funds because of disease or for treatment are easy to obtain donations from potential donors [30].

By investigating 13 studies, the key factors of successful medical crowdfunding campaigns were extracted. On this basis, we conducted a more in-depth review and provide a comprehensive understanding of the factors that influence successful campaigns. We find the success factors of medical crowdfunding campaigns could be divided into 4 categories: platforms, raisers, donors, and campaigns. The success factors involve the main actors in the whole medical crowdfunding campaigns, but campaign factors were more frequently studied than other factors. Platform factors, however, played an important role in medical crowdfunding [47], because platforms are important in linking donors and raisers [48]. There were relatively few studies related to donors; thus, future studies could pay more attention to this area. Despite the widespread use of medical crowdfunding, to the best of our knowledge, this is the first systematic review that examines the factors that lead to successful medical crowdfunding campaigns.

The 4 categories of successful factors that are actors in medical crowdfunding interact with each other [7,49,50]. First, raisers register in the platforms to initiate campaigns to raise money for medical expenses with illness narratives, which include fundraising goals, fundraising time, and fundraising events. Second, the medical crowdfunding platforms conduct a preliminary review to release the campaigns. Third, managers of the platforms give the funds raised from the public to the campaigners after the relevant processing charges are collected. A crowdfunding campaign will be success if all the above steps are were successfully implemented.

Theory for understanding the success factors of medical crowdfunding can be developed from our findings. None of articles included in this review take strong theoretical perspective to help analyze, explain, and predict the success factors of medical crowdfunding and convey few theoretical implications. Therefore, it is necessary to apply more theories to better understand this topic. Second, more methods or mixed methods could be employed to investigate this topic. Although text analysis was used most often, more insights could be explored using other methods that possess different advantages. We found that data were collected from single sources, which may result inherent biases; collecting data from multiple sources could allow validation to proposed research models and used methods in different studies.

Our study has some limitations. First, no new factors and relationships about medical crowdfunding could be explored through systematic review. This was the most significant inherent limitation of our study since we relied on previously published literature. Second, additional literature could have been included in our systematic review. Literature written in other languages from other databases may also have discussed the success factors of medical crowdfunding. Third, bias may still exist in the process of selection and evaluation of literature. Although we attempted to minimize bias from differences between reviewers, such as solving conflicts in judgment by discussion among reviewers, bias cannot be excluded. Fourth, heterogeneity exists in the literature included in the review; therefore, the quality of our review results may be impacted. Other categories of success factors, such as health systems or national economic status, could be also be considered.

By examining platform, raiser, donor, and campaign–related success factors, we provide information that can be used as the basis for future research and future medical crowdfunding campaigns.