We all want to live at home as we age. But while we may not want to admit it, getting the support we need can be much tougher at home than in a nursing home or assisted living facility.

For all their problems, nursing homes or residential care facilities can be an efficient way to deliver care to many people who live in one building or on one campus. Aides can visit multiple residents by walking down the hall. A community bus can provide transportation. Social supports and activities are readily at hand. Meals are available in the dining room.

Compare that to people living in their homes. Yes, they live more independently and in familiar surroundings. And they don’t have to make their day fit with a facility’s schedule. But bringing them the assistance they need can be complicated and expensive. And it often suffers from a complete breakdown of coordination. How does an older adult living at home learn what services are available? And even if she knows, how does she make sure she gets help when she needs or wants it?

The answers to these questions are not easy. Yet, until we answer them, living at home may not be the solution that many of us hope for. It can be unsatisfying and even dangerous.

Fortunately, several new housing models are experimenting with ways to better deliver care to seniors living at home. And we are just beginning to learn how they are working out.

One closely watched program is a Vermont initiative called Support and Services at Home (SASH). The program is a partnership between a non-profit developer and local service providers such as home health agencies or councils on aging. The idea: Bring care coordination and some basic services to older adults living in or near low-income senior housing.  The state has also agreed to make advanced primary care available to those using the program.

Participants are all enrolled in Medicare, which pays the initiative a per-member per-month fee to operate the program. About 2,000 seniors participate—about 1,500 residents and about 500 neighbors.

Last year, RTI International and Leading Age (a group that represents many non-profit providers) evaluated the first year of SASH, and the results were promising but far from definitive. They found the model reduced total Medicare spending by between about $1,700 and $2,200 per participant, compared to similar seniors who did not participate in the program. However, the evaluation also found that SASH participants used more hospital services than those who were not in the program.

Last month, Leading Age and the consulting firm the Lewin Group completed a separate study of SASH and several similar “housing with services” programs that offer on-site service coordinators to help residents plan and organize care.

That study showed lower hospitalization rates but more visits to the emergency department that did not result in a hospitalization. Keep in mind that this study only looked at sites where these services were offered. It did not have good information about whether residents actually used them. Like the first study, this one suggests that better services available through senior housing may reduce medical costs, but is far from definitive.

In a recent paper, Prabhjot Singh of the Mt Sinai Health System in New York and Stuart Butler of the Brookings Institution looked at some of the challenges and opportunities of using health systems as hubs for this sort of care integration. The most successful found ways to work with networks of local social service providers to better integrate medical and social care. But cultural and financial issues often got in the way.

Increasingly, medical and social service providers, government, and advocates are coming to appreciate the importance of better coordinating care for people with chronic conditions. Progress has been slow, in part due to mix of distrust, financial disincentives, bureaucratic hesitation, and an absence of clear evidence of that these models improve care and save money. But we are starting to learn more about what does, and does not work.