Table 6.

Recommendations.

#	Recommendation	Description
1.	Integrate arts/culture opportunities into existing referral networks.	Referring patients to non-medical resources and experiences is not new in healthcare; however, the presence of arts/culture among those referral options is new. Given this evaluation's findings, we recommend adding such opportunities to existing community-referral networks. Doing so is likely to be welcomed by healthcare providers and other community organizations (see Recommendation 7), and it will increase options for health promotion.
2.	Put more funds toward creating robust processes for building and promoting the program.	Grant funds were initially perceived as necessary primarily for reimbursing the cost of the free services prescribes receive. However, reimbursement is only necessary if referrals are made and used. Thus while a program builds, it is critical that funds be allocated to program-building processes such as cultivating and stabilizing partnerships with healthcare providers, promoting the program (to providers, other community agencies, the public), developing robust processes for participant connection and follow-up, ensuring data is collected and documented, addressing access concerns, etc.
3.	Expand the number of healthcare providers partnered with each organization.	The low number of referrals to some organizations appears related to the limited number of healthcare providers with which they were partnered. Cultural organizations have the capacity to be partnered with multiple providers, as they are unlikely to receive too many referrals. By contrast, too few may significantly limit the program's reach. To address this, cultural organizations could add provider-partners individually, or co-develop healthcare partners with other organizations. As with MACONY/CHC/UCP, being partnered with multiple cultural organizations gives providers and their patients more options; similarly, multiple provider-partners ensures more engagement for each cultural organization.
4.	Expand the number of participating cultural organizations.	Many providers requested additional options for patients/clients. A starting point would be to encourage current organizations to work together, since they would then share provider-partners (This is currently modeled by the five cultural organizations partnered with MACONY/CHC/UHP). In addition, given concerns about equity and access–including relevance and cultural responsiveness–we recommend ensuring that smaller, grassroots arts organizations are sought for participation. Such organizations may have significant meaning for specific communities, yet may not have ready access to typical grant opportunities.
5.	Design a website and one-pager to help healthcare providers quickly recognize the varied health benefits of each program.	Providers found CultureRx beneficial, but demonstrated a limited understanding of the varied ways in which referred opportunities could support patients/clients. In response, cultural organizations should generate a concise, research-based summary of the ways in which their program(s) may benefit participants (As one example, see Canada's PaRx program website). Funds should be set aside to both research these program-benefit descriptions and make them readily accessible.
6.	Consider alternative activities and schedules.	Some participants do better with fixed-schedule events such as classes, while others respond well when they can attend whenever they would like. Whichever scenario is most common for a given organization, we recommend they imagine how they might complement it by occasionally incorporating an alternative option.
7.	Pilot the use of digital platforms to link cultural organizations with healthcare providers and other social/community resources.	A growing variety of digital community-referral platforms are supporting community health by providing structures for referrals and community engagement. Many utilize screening processes related to social determinants of health, identify relevant potential supports, and then provide tracked referrals to services (e.g., to transportation, mental health, shelters, etc.). Reports from these platforms indicate parallels with this evaluation's findings, such as (1) the need for extensive patient followup, (2) the need for providers to learn more about community resources and how to use them, and (3) the well-being benefits providers receive from being able to offer non-pharmacological solutions (56). Such parallels suggest that arts/culture assets could be readily integrated with digital community-referral platforms to support health and healthcare. In addition, most platforms do not explicitly include arts/culture in their networks; as a result, piloting such inclusion would provide new, useful information regarding community health practices.
8.	Collect data from all participants, rather than strictly those being referred/prescribed.	Data collection for this study focused on CultureRx participants. However, the benefits or challenges of cultural participation are not limited to individuals receiving a healthcare-based referral. In the future, collecting data from all participants will generate more information regarding health outcomes–thus better informing providers' decisions to recommend an experience. In addition, asking everyone to provide feedback prevents singling out those who are present due to a referral, and may also bolster public awareness of the initiative. Of course, data collection should include a way to note how a given respondent heard about the program, so that data from referred participants can be extracted for analysis if necessary.
9.	Share data collection successes and tips.	Some CultureRx organizations learned successful processes for data collection, which improved participants' experience. Given that such learnings are likely transferable across organizations, they should be regularly shared. Similarly, when organizations encounter difficulties or unusual situations, they should be given opportunities to walk through their process with a program evaluator or, perhaps most helpfully, with similar organizations that may have helpful tips. As mentioned, organizations should also be urged and expected to use funds to enhance referral and data collection processes. Finally, as more arts-on-prescription programs are created and studied, evaluation methods and practices should be shared and analyzed to support improvement and synthesis.
10.	Address structural barriers to equity, access, and inclusion.	Initiatives such as CultureRx face equity challenges common to many other sectors and systems, including sustainability (short-term funding that expects long-term outcomes); culture (knowledgeability and representation regarding diverse communities served); and selection bias (organizations with the most access to competitive grants may not be the organizations most prioritized by communities). These concerns require intentional inquiry and action, as suggested by the following questions [quoted from Golden et al. (36)] • Is the organization or its programs in (or accessible to) marginalized communities? • What relationships do they have with such communities? • How is the voice of these communities informing the design and evaluation of the organization's programming? • How does the leadership's demographic make-up reflect that of the communities they are trying to reach? • How does the mission of participating programs and providers align with that of the overarching initiative?
11.	Implement frameworks for becoming antiracist and inclusive.	Several tools and frameworks exist for use by health service organizations to self-assess and prioritize areas for improvement on a continual basis [see (57–59)]. These standards and tools can also be used by arts/culture organizations to support them in increasing inclusivity of–and effectiveness for–diverse and marginalized communities.