“This is one way of demonstrating through the data that you have a more medically complex person,” says Barbara DiPietro of the National Health Care for the Homeless Council.
When Lehigh Valley Health Network (LVHN) determines that admitted patients are homeless, physician assistant Brett Feldman, director of the eastern Pennsylvania system’s Street Medicine program, gets a call.
“They will request a consult for Street Medicine in the hospital, just like they do for any other specialty service,” Feldman says. “We don’t take over care, but we walk the inpatient team through what this person’s version of homelessness looks like and how that might be impacting the patient’s health.”
The team provides approximately 20- 25 consults each month. Because the team knows patients’ living conditions, they can recommend treatment plans and discharge instructions that are likely to be successful. In 2014—the program’s first year—the 30-day readmission rate for homeless patients with a Street Medicine consult was approximately 50 percent. That fell to 31 percent in 2015 and to 13 percent in 2016.
As health systems start to recognize the connection between homelessness, poor health status, and the potentially avoidable use of emergency and inpatient care, they need to document homelessness among their patients, says Barbara DiPietro, senior director of policy for the National Health Care for the Homeless Council (NHCHC). Doing so is likely to improve care outcomes and validate the high-acuity level associated with homeless patients.
“This is one way of demonstrating through the data that you have a more medically complex person,” she says.
Coding for Homelessness
NHCHC is a membership association made up of organizations that receive federal Health Care for the Homeless grants, medical respite providers, and other stakeholders working to address the healthcare implications of homelessness. In a policy brief issued last fall, the council encouraged health system leaders to consider using the Z59.0 code, an ICD-10-CM code to document a patient’s homelessness (DiPietro, B., and Edgington, S., Ask & Code: Documenting Homelessness Throughout the Health Care System , National Health Care for the Homeless Council, August 2016).
Federal guidelines specify that the code can be used in any healthcare setting and may be used as a principal diagnosis code in the inpatient setting or a secondary code, depending on the circumstances.
The NHCHC cites several reasons for using Z59.0 to document homeless patients:
- The ability to compare data across provider organizations
- The opportunity to signal patients’ high-risk acuity to various providers when patients and their medical records move within and across systems
- The ability to document the medical complexity of a system’s patient population compared with that of other providers
- A way to track medical outcomes of a system’s homeless patients
- A reduced chance that providers will be financially penalized under value-based payment methods for serving high-needs populations
As provider organizations become aware of the Z59.0 code, they are enthusiastic about its potential, DiPietro says. However, their interest has not translated into rapid adoption of its use.
One reason for the slow adoption is that most provider organizations do not screen for homelessness and, therefore, may not be aware of—or certain about—patients’ housing statuses. Beyond that, there is no payment associated with a homelessness code, so training providers and coders to use Z59.0 tends to be a lower priority than other coding issues.
However, as value-based payment methods take hold, DiPietro predicts the code will be embraced as a way for providers to document that they are caring for particularly high-acuity populations.
“We know where we need to go in terms of value-based payments that are adjusted by acuity, but there are very few systems that have moved to actually being paid for outcomes,” she says. “We are in an in-between space now.”
See related sidebar: Yale New Haven Hospital Homeless Patient Screening Tool
Screening for Homelessness
Of course, coding for homelessness requires screening, which is still uncommon for most provider organizations. While many screening tools exist, none has become standard or identified as a best practice, DiPietro says. One thing that doesn’t work: Directly asking “Are you homeless?”
NHCHC’s policy brief explains why:
- People without homes may consider their temporary or tenuous living arrangements to be “homes.”
- They may not want to be labeled homeless because of stigma or shame.
- They may provide friends’ or relatives’ addresses or previous addresses.
- They may fear that being identified as homeless will lead to poor treatment.
Homeless Screening in Practice
LVHN, based in Allentown, Pa., has operated its Street Medicine program for four years. The program is staffed by Feldman, a nurse, and approximately 200 volunteers, of which about 80 are clinicians. The team treats patients at homeless shelter clinics and on the streets. If patients present at an LVHN emergency department (ED), the EHR identifies the patient as homeless, triggering a consult with Feldman’s team. In addition, patients who have not been served by the Street Medicine program but are identified at the hospital as homeless may receive a consult if the ED physician or hospitalist wants assistance.
Suspecting that the Street Medicine team was seeing only a small portion of the homeless patients treated at LVHN, Feldman conducted a study of the health system’s three hospitals to determine what percentage of ED patients were homeless. A total of 4,395 patients—a sample determined by the study design—were screened over a six-month period in 2015 and 2016.
The finding: 14 percent of patients presenting at the inner-city hospital were homeless, as were 6 percent of patients at one suburban hospital and 5 percent at another suburban hospital. Overall, the homelessness rate was 7 percent across the three EDs.
That means that the hospital with the biggest ED—45 beds—and the lowest rate of homelessness—5 percent—on average has at least two beds occupied by homeless patients at any point in time.
“There were many more homeless people coming into our EDs than we realized, which means that there are a lot more people using the healthcare system in an inefficient way, with higher readmission rates and higher length of stays than we realized,” Feldman says.
The Benefits—and Challenges—Of Knowing
Feldman identified two benefits for health systems that routinely screen for homelessness in their EDs, so that patients’ housing statuses are known by the time they are admitted.
Shorter length of stay (LOS) for homeless inpatients. Feldman was prompted to conduct his study because he noticed that average homeless patient LOS was nine days, compared with five days for housed patients, even if Street Medicine was involved.
As he studied the consult patterns, he realized that the Street Medicine consults typically did not occur until five days after admission when hospital staff were beginning to wonder about effective discharge plans. “We had already missed out on being able to provide recommendations on the treatment plan because that has been determined by then,” he says. “It is impossible to make an impact on the length-of-stay if we’re not even coming on board until that late.”
The opportunity to find payer sources for homeless patients. One of the Street Medicine program’s top priorities is securing Medicaid coverage for its patients. That helps providers, and it encourages patients to use preventive and primary care services.
“When they have insurance, it empowers them because when they need to use the traditional healthcare system, they are not a charity case,” Feldman says. “They have something they can contribute in exchange for what is being provided for them.”
Typically, Street Medicine staff and volunteers devote a lot of time to encouraging homeless individuals to apply for Medicaid, arranging meetings to fill out paperwork, securing birth certificates, and assisting with other steps in the process, all of which is laborious because homeless people can be difficult to find or reach by telephone.
“When they’re in the hospital, that’s a golden opportunity for us to get as much done as possible because they’re a captive audience,” Feldman says.
Although the LVHN study did identify the homelessness rate in EDs, the health system did not implement a comprehensive screening program going forward. That’s because only two clinicians—Feldman and a nurse colleague—can do the inpatient consults, and they are already responsible for 2,000 homeless patients, and Feldman manages the program’s administration.
Thus, the 20 to 25 consults a month that they currently receive are all they can realistically handle. “Once we start screening everyone, we know the volume of consults is going to increase, and we don’t want to take that on until we can handle it,” he says.
An Opportunity, Not a Hassle
Most health systems are just beginning to examine the impact of homelessness and other social determinants of health on their patient outcomes and operational performance. While all health leaders understand the importance of data, DiPietro recognizes that additional screening questions and data collection may seem overwhelming.
She hopes healthcare leaders can frame the emerging awareness about social determinants as an opportunity to get at the root causes of heavy healthcare utilization—an area where revenue cycle, patient access, and coding staff can play a role by screening and documenting homeless patients. After all, the social determinants—homelessness, domestic violence, hunger, veteran status, addiction, and others—have affected health systems for decades, but they generally have not been quantified and analyzed so they can be effectively addressed.
“We are just starting to turn the curve on our awareness on how this impacts our costs and our outcomes,” DiPietro says. “I hope healthcare leaders can get excited about the opportunities that this presents for us to provide better care, rather than feeling drained by what can seem like ‘piling on.’”
Over time, screening tools and an ICD-10-CM code that allow providers to identify and document homelessness will benefit patients and healthcare providers alike.
Read more about LVHN’s Street Medcine program in HFMA’s Strategic Financial Management newsletter at hfma.org/sfp/StreetMedicine.
Or listen to an interview with Brett Feldman of LVHN from HFMA’s Voices in Healthcare Financie Podcast at soundcloud.com/hfma/voices-in-healthcare-finance-episode-7.
Interviewed for this article:
Barbara DiPietro is senior director of policy, National Health Care for the Homeless Council, Baltimore.
Brett Feldman is director of the Street Medicine program, Lehigh Valley Health Network.