Changing Client Outcomes
Or, why data standards consensus is critical
Well, it’s not every day a social worker goes to a data standards conference. Last week’s conference was a foray into a foreign, jargon-filled world, yet it was designed to create a data standard that would allow social workers (and others!) to share aggregated information about all services available to people in need.
The problem is fragmented data. Each nonprofit keeps their own records about what services are available in the community, and consequently, the data is in varying stages of accuracy and completion. There is either too little information, too much information (and poorly organized), or just plain wrong information. Social workers use this data about the services in the community to send clients to other nonprofits, also known as making a “referral”. When the data is bad, the referral is bad.
When I arrived at the conference, I thought I was the only one irritated with the state of data quality in referrals, but suddenly, I was immersed in a room of socially conscious people who cared about how data can influence client outcomes. I was in heaven.
As a social worker, first at Catholic Charities and now at Tenderloin Housing Clinic, I have seen how clients are oftentimes unable to get the help they need, in the timeline that they need it, due to the difficulty in accessing accurate information. A project called Open Referral, sponsored by Code for America, would help individual agencies share their resource directories with each other and with the big resource aggregators like 211, meaning that everyone in the social services ecosystem would have access to better referral information (can I get an amen?). Eventually, search engines like Google would be able to serve up the relevant data without social workers having to dig through individual web pages on each nonprofit’s (generally outdated) website.
So, who was in the room? There were nonprofits, social workers (actually, just me!), social services businesses (like Purple Binder), data administrators, and community organizers. There was also a contingent of folks from Code for America — the Pied Pipers of open civic data. They birthed the Ohana API (aka application programming interface, which is just a fancy term for how computer systems talk to each other), a project to take social services data from San Mateo County and open it up to the public so service providers and individuals looking for help could easily find it. This (but nationwide!) is what Open Referral is trying to achieve. Hopefully other cities follow in their footsteps and share their data too.
As we kicked off the conference, participants role-played how the current system fails clients in crisis by providing referrals that don’t fulfill the client’s needs or don’t come quickly enough. The skit depicted a woman at home with two children under five. She had been previously battered by her partner, who was getting out of jail that night. The woman was afraid of what would happen when he came home, and wanted to find a domestic violence shelter for herself and her two children that night, and pronto.
The client called an information and referral (I&R) agency worker who put the client on hold, while he called another social worker that could confirm which domestic violence shelters accepted children under five and had three beds available. The social worker looked in a resource binder that that had not been updated, so she had to call around to find out which domestic violence shelters had beds available. The time it took to get the relevant information was time that was trickling away until the batterer returned.
Ultimately, a domestic violence shelter bed offer is only useful if the woman is able to leave her home quickly and safely, if the shelter accepts young children, and if the shelter has enough beds for the three of them, and that information needs to be at the ready at all times. While the skit wasn’t totally accurate — not one I&R worker or social worker turned to Google (basically a social worker’s best friend) — the heart of it still rung true.
When time is of the essence in referrals, having the data online is generally better than having to call for it. Not only is it quicker, but there are certain things that you just don’t want to have to talk about in person or over the phone. The importance of being able to access relevant information online was illustrated by the United Way, who told us that they see more people searching for STD services anonymously on their website than they have contacting their call center. Surprise, surprise.
There tends to be some resistance from nonprofits in the push towards getting data online, probably due to the misguided belief that clients (such as the ones profiled here) do not have the digital skills to search for their own help online. Don’t get me wrong: I understand that 38% of my formerly homeless single adult clients don’t have regular access to the internet, but I also know that 77% of the 135 surveyed tenants at the Seneca Hotel stated that they wanted wifi in the hotel. The desire and the willingness to learn is there, even if the skills to create an email address aren’t. And the skills can be taught.
After we explored the failings of the current system, we parsed through the different use cases and came up with four main users. They were:
Help Seekers: these are the clients I deal with daily — clients who have experienced some sort of trauma, and has difficult processing next steps or how to move forward. They face stigma for seeking help, and do not have access to the information they need in a prompt manner.
Help Providers: That’s me! The social workers, case managers, direct service providers, community health/public health workers, EMTs, and patient navigators of the world. Also, Information and Referral (I&R) agencies like 211 and Eden I&R, which are organizations that the public can call and request a referral to an agency that can help them.
Researchers: These are the ones who use the data for policy implications — the people who want to make sure that we are making evidence-based decisions when implementing social services programs. They also want to know things like the number of available nursing beds throughout the state, so if an earthquake hits, they can give the city reliable and accurate data to relocate medically fragile patients.
Data Administrators: these are the folks who are managing the quality of the data. They are motivated by getting accurate information captured and entered. Thank goodness for these people: I’d lose my mind doing their jobs, though I’m sure the feeling is mutual.
Once we had figured out who exactly we were targeting, we arrived at a couple of hypotheses that Open Referral plans to test: First, that it is possible to build and maintain a distributed resource data system in which user-submitted input is reliably validated (exactly how that will happen is what we’re still figuring out!). And second, that this system could be governed by a set of local stakeholders, in partnership with a nationwide governing body that would steward the new standard.
So, let’s change the (social services referral) world! Learn more about the history, check out the FAQs, comment on the specifications, or just tweet the organizer to learn more.
It may take a while, but the times are changing! I’m hoping the social workers of the future won’t have to struggle to keep up with the moving targets of accurate social services information. Because, really: Problems have solutions. And this is one we can definitely solve.
