Homeless deaths in L.A. County are leveling off but still nearly seven per day

Los Angeles County health officials reported a second consecutive year of moderate increases in homeless deaths, adding evidence that a surge that raised the mortality rate 56% over the previous two years has leveled off.

But homeless mortality remained persistently high with an average of 6.9 deaths daily in 2023, the year covered in the most recent report. Overall, homeless people died at 4.5 times the rate of the whole population.

During the year, 2,508 people died on the streets and in shelters, up 5.6% over the prior year. Because that period’s homeless count increased, the mortality rate — calculated as 3,326 deaths per 100,000 people — increased less, at 1.3%. That was an improvement over the 2.1% increase in 2022.

The rate of drug and alcohol overdose deaths fell slightly but remained the leading cause, accounting for 45% of all deaths. Of those, 70% involved the synthetic opioid fentanyl. Coronary disease was the second leading cause, accounting for 14% of deaths, a figure that was up by 22%. Transportation accidents remained steady in third place overall, but the second leading cause for women. On average, one homeless person was killed in a vehicle accident every other day. Homicides, in fourth place, fell 25%.

Men were more likely than women to die from all causes, making up 68% of the population but 82% of deaths. Sixty-two percent of deaths were among those younger than 55, a phenomenon attributed to the younger group’s higher overdose rate. Mortality was highest for whites who were 19% of the population but accounted for 32% of all deaths. Black people and Latinos each had fewer deaths than their proportion of the population.

At a news briefing Thursday, Dr. Gary Tsai, county director of the Bureau of Substance Abuse Prevention and Control, said additional contributing factors for the leveling off could be growing awareness within the homeless population of the risks of fentanyl and even the diminished number of fentanyl users due to deaths.

But he noted that there have been no sharp changes that might indicate those would be significant factors.

“We know that people who are newly homeless can be at increased risk for overdose,” Tsai said.

The sixth report on homeless mortality by the Los Angeles County Department of Public Health described the current trend as a plateau. It attributed the improvement partly to a threefold increase in the distribution of naxalone, an overdose reversal medication, from 2021 through 2023 and other mental health and substance-use treatment services. Nearly 479,000 doses of naxalone were distributed in 2024, it said.

The report listed 17 recommendations falling under four categories: rapid access to housing and shelter; expanded overdose prevention; physical, mental health and substance use treatment; and collaboration with municipalities and unincorporated communities to reduce traffic deaths.

The calculation of homeless mortality rates is inherently unstable. It relies on fluctuating annual counts of the homeless population that have built-in error and on imperfect means of identifying deaths of homeless people.

The deaths are drawn primarily from the Los Angeles County Medical Examiner’s records of accidental and violent cases. Other deaths are obtained from a search of the state database of death certificates.

The year-over-year increase in 2023 may have been inflated by an improvement in the state data with the addition of a homeless checkbox. State data contributed 12% of the identified deaths, more than in any previous year.

This year’s report also included a footnote indicating that the initial 2022 finding of a leveling off of the overdose rate had to be revised upward due to a backlog of toxicology tests at the time of publication that the medical examiner later determined to be drug-related. A smaller backlog also could change the 2023 result, but to a lesser extent, it said.

The homeless counts are averaged over two years to estimate a mid-year number. This year’s report used a three-year average due to the cancellation of the 2021 count during the COVID-19 pandemic.