Code Z59.0: diagnosing the toll of homelessness on health care

Open this photo in gallery:

In the emergency department at the Royal Alexandra Hospital in Edmonton, Dr. Louis Hugo Francescutti recently treated a patient who was homeless and was there for the 360th time.

“I’ve had others that are in the 500 range,” he said in a recent phone interview from the hospital. “I’m at work right now – I came in at 6 o’clock, and the first six patients I saw were experiencing homelessness.”

With homelessness at crisis levels in cities across the country, emergency-room doctors and nurses have become a frequent point of contact for people without shelter, who face elevated risks of injury, illness and death and often have nowhere else to turn.

The medical toll of homelessness has become so acute that hospitals are now required by the Canadian Institute for Health Information (CIHI) to track whenever a lack of shelter is mentioned in a patient’s chart.

But doctors and nurses are not obligated to ask patients about their housing status, so this information is not always making it onto those charts in the first place – meaning the picture is incomplete.

And without data, the toll of the crisis remains difficult to measure and address – and opportunities for intervention are being missed.

Under the World Health Organization’s International Classification of Diseases (ICD), which is used to track global health statistics, homelessness is coded as Z59.0.

In 2018, CIHI, in consultation with provinces and territories, mandated that this code be used by administrative data-keepers anytime homelessness is referenced in a patient chart. So when they see “no fixed address,” “patient living in shelter” or “homeless,” they input Z59.0 – the same way U07.1 is entered for a COVID-19 diagnosis or S82.2 for a tibia shaft fracture.

“Housing status, and in particular homelessness, is a huge determinant of health. It has a huge impact on people’s health, and a huge impact on their use of health services,” Keith Denny, CIHI’s director of Population and Indigenous Health, and Classifications and Terminologies, said.

In those four years since it was mandated, Mr. Denny said, CIHI has seen an improvement in the capture of the data. But it’s far from perfect, and varies from hospital to hospital, and from province to province.

“The staff who are responsible for assigning the ICD codes can only use Z59.0 if homelessness is noted in the patient record. We have no way of knowing how many people experiencing homelessness are not identified as such,” Mr. Denny said.

One of the biggest opportunities for improvements in the Z59.0 data collection is in emergency departments – the part of the health care system that homeless patients most frequently access

The nature of emergency care – focused and brief interactions – means that housing status is less likely to get logged in those cases.

“When you admit someone to hospital, you do take a fairly complete history and describe where they live and who’s living with them,” said Dr. Stephen Hwang, a general internist and director of the MAP Centre for Urban Health Solutions at St. Michael’s Hospital in Toronto and one of the world’s leading researchers on homelessness, housing and health. “But you don’t do that in the emergency department, necessarily.”

CIHI plans to publish its first report on Z59.0 data this year, with a goal of raising awareness of the code for both physicians and researchers across the country, to get them thinking about the value of asking the question, both clinically and statistically.

“We feel there’s a real opportunity here to get better data,” Mr. Denny said.

The U.S. Centers for Medicaid and Medicare has been at the forefront of such data collection, and by next year will make it mandatory for all hospitals reporting to its Inpatient Quality Reporting (IQR) program to submit information on social drivers of health, such as housing, as well as the number of patients who were screened for these factors.

One of the primary challenges with a code like Z59.0 is that homelessness is nuanced, and there may be conflicting ideas about how it should be defined. Many people think of homelessness as someone sleeping on the street or in a shelter, but there is also hidden homelessness, such as people couch-surfing or living in inadequate housing.

Dr. Hwang says social factors play a huge role in a patient’s health, and are critical to assess on a systemic level.

What the Z59.0 data provide, he says, is “the opportunity to understand the connection between homelessness and use of hospital care – because we know that people who are homeless are more likely to be admitted to the hospital. We know that their hospitalizations can be prolonged, or more complex, and also that they’re at higher risk of readmission afterwards.”

Dr. Hwang encourages any effort to bolster the Z59.0 data – particularly to set benchmarks, so that progress can be measured moving forward. At the same time, he stresses, the problem is already clear.

“I think it’s more important to take action to improve the situation,” he said. “We have enough data to know that we have a long ways to go.”

At a time when hospitals and emergency departments are under so much strain, Dr. Francescutti agrees that it is critical to address homelessness as a health crisis now.

“The most expensive part of the system is wasting millions of dollars when these folks would be far better served in a different environment,” he said.

Just this week, Alberta announced a pilot project that will dedicate 36 transitional beds for people who would otherwise be discharged from Edmonton emergency departments into homelessness.

Dr. Francescutti said he believes all patients should be asked about their housing status in a standardized way when they arrive, either by a triage nurse or registration clerk. Asking for an address, which is already standard, he said, is not enough, because some people will provide a relative’s or friend’s address or a shelter address.

“The question is really very simple – at the very beginning, you ask in a non-intimidating way: Are you experiencing homelessness or are you sleeping rough? And if they say yes, then that gets coded.”

According to an Alberta Health report on Z59.0 data – which Dr. Francescutti stresses is an undercount – there were 26,396 visits to Alberta emergency departments and urgent care centres by people experiencing homelessness last year. In 2020, there were 24,926. In 2019, there were 24,750.

The CIHI report will offer similar data, but countrywide.

“We know that those numbers are underrepresented,” he said. “But in spite of that, it tells you that there’s enough of a problem that somebody should be pushing the panic button.”

Like the patient he saw recently who was there for the 360th time, Dr. Francescutti says many of those visits are repeat visits by people who are then cycled back into the same illness-inducing circumstances that led them to attend in the first place.

“Today in Edmonton, it’s -32 C,” he said, on a recent Friday morning.

“And you can rest assured that if I can’t feel comfortable that a patient has a safe place to go, they’re not going anywhere – I’m going to keep them. And if I keep them, that means I’ve got one less bed for other patients in the waiting room.”