It is difficult to tease out how much naloxone has contributed to a reduction in overdose death rates.
In some states, including Ohio, Washington and Oregon, methamphetamine was also closely tied to overdose death rates, according to the Guardian’s analysis.
Catoosa county, Georgia, saw an 120% increase in overdose deaths since the national peak, the highest among counties with data available in the Guardian’s analysis.
The Guardian’s analysis of Millennium Health data found that the more you zoom in, the more specific trends emerge.
Portugal’s per capita overdose death rate is less than 1/30 of that in the US.
According to an exclusive data analysis by the Guardian, overdose deaths in the US decreased nationally in 2024, but they increased in some communities. The analysis also revealed significant regional differences in the number of deaths associated with the public health emergency.
The announcement follows months after public health officials announced a 27 percent decrease in overdose deaths, which the Centers for Disease Control and Prevention (CDC) credited to changes in the drug supply, increased access to substance-use treatment, and the overdose-reversal medication naloxone.
“This decline suggests more than 81 lives saved every day,” the CDC stated in a press release. In 2024, an average of 220 US citizens perished from overdoses every day.
But according to the Guardian’s findings, the downward trend in overdose deaths had already started to reverse in seven states when the CDC announced the decline in May 2025.
Experts in public health were uncertain about the causes of the drop and whether the drop in fatalities was the start of a longer-term pattern. Five experts can be consulted, and each one will offer a different explanation for the decline in deaths. According to Andrew Kolodny, medical director of Brandeis University’s opioid policy research collaborative, “nobody really knows.”.
The Guardian analyzed important data points, such as CDC data on overdose fatalities and the distribution of medication for opioid overdose and addiction; US Census Bureau data on population, poverty, Medicaid, and uninsured rates; and drug seizure data from the National Forensic Laboratory Information System, to better understand how—and why—the overdose epidemic is evidently declining in some regions of the nation while progress is waning in others. Data from the urinalysis business Millennium Health, an authorized drug-testing specialty laboratory, on patients undergoing substance-use treatment was also examined in the analysis.
In some US counties, overdoses increased by as much as 120 percent between August 2023 and December 2024, the most recent month for which county-level data is available, even though overdoses decreased nationally during that time. According to state-level data that is available through April of this year, deaths started to increase once more in January. They approached peak levels in Arizona and increased more slowly in Delaware, Hawaii, Minnesota, New Mexico, and Washington.
According to the CDC, overdoses continue to be the leading cause of death for Americans between the ages of 18 and 44. If they weren’t grouped together with other unintentional deaths and injuries, they would rank among the top 10 causes of death overall.
More than 80,000 people in the US lost their lives to overdoses in 2024, which is more than twice as many as car crashes and almost three times as many as COVID-19. However, this was still considered a significant advancement.
The lack of national focus on addiction treatment frustrates Marc Fishman, an addiction psychiatrist at Johns Hopkins University who specializes in substance-use disorders in young people.
“Who knew we’d be cheering 80,000 deaths?” asked Fishman. “There would be cause for celebration if 80 went to 70, which would go to 60. We’ll see if the pattern persists. “.”.
pharmaceutical supply.
Among the variables examined by the Guardian, variations in the quantity of fentanyl and, occasionally, methamphetamine in the drug supply, along with geographic location, were most closely associated with shifts in overdose fatalities.
Recent research and the Guardian’s analysis suggest that these two factors are probably related. A September study published in the Lancet states that “fentanyl started replacing heroin in the eastern US and spread westward, driving profound increases in overdose deaths” that haven’t reduced as much as they have in the east since 2013. Many Americans developed an addiction to opioids as a result of doctors prescribing more of them starting in the 1990s, particularly in eastern states like West Virginia where physical laborers were using pills to treat their pain. An increase in the use of illegal opioids, like heroin, was caused by stricter regulations on prescription opioids that began in the early 2010s. It was then gradually replaced by fentanyl, which is 70 times more potent and less expensive.
According to the Guardian’s analysis, peaks in fentanyl prevalence coincided with peaks in overdose death rates, and 8% of people in western counties included in the analysis lived in areas where overdoses were still rising by over 10% since the crisis’ peak. In 2023, the year that overdoses started to fall nationally, the percentage of drug seizures that contained fentanyl peaked in 28 states.
Naloxon.
In CDC releases, the overdose-reversal medication naloxone is frequently mentioned first as a potential reason for the drop in drug overdose deaths, and it is widely hailed as a game-changer in the crisis.
When the body forgets to breathe due to an overload of opioid receptors in the brain, an opioid overdose occurs. This process is reversed by naloxone, which blocks the brain’s opioid receptors and causes the recipient to go into withdrawal as soon as they wake up.
“It is terrifying,” Fishman said. For years, I have been screaming about this. Without addressing the disorder, you cannot declare that you will reverse an acute overdose and then rejoice. ” Substance-use treatment can help prevent overdoses in the longer term, but it requires sustained investment in both addiction and mental health treatment. Those who battle addiction frequently use self-medication to treat other mental illnesses like PTSD or depression.
It is challenging to determine the exact amount that naloxone has helped lower the number of overdose deaths. The available data is conflicting, and the few studies that look into how naloxone affects overdose rates were primarily carried out prior to the sharp increase in overdoses following COVID-19.
blindly flying.
As fentanyl levels in the drug supply continue to rise, the CDC released recommendations in 2015. There have been some, like an increase in naloxone distribution. Numerous others haven’t, such as regular examinations of medications discovered at overdose scenes.
In reports about fewer overdose deaths, the CDC praised its enhanced data sharing and collection. Experts say it is not enough.
Kolodny stated, “We ought to have much better data on this.”. “You could find out about Covid positivity rates on tests if you think back to when the virus was rampant, on a county and zip code level. Everyone had easy access to information about hospitalizations and deaths. Almost thirty years into the opioid epidemic, we still lack sufficient surveillance. “.”.
CDC overdose death data is “the best we have” to monitor the crisis, Kolodny said, adding that it is “inadequate as it is.”. We are examining [April data]. It’s absurd. October is here. “”.
Millennium Health is already collaborating with certain health departments to “close the gap between what’s happening with fatal overdose today and what we’re being told today happened six, seven months ago,” according to Eric Dawson, vice-president of clinical affairs. According to a modeling study that was published in Jama this year, overdose deaths could be reliably predicted using urinalysis data from Millennium Health’s patients undergoing drug abuse treatment.
Spikes cannot be identified because the CDC only publishes data for the previous 12 months. There is data on the demographics of the fatalities and the type of drug that caused the death, but it is very limited; for instance, it is impossible to determine the number of deaths that were caused by multiple drugs.
According to new Trump administration policies, the CDC may never release data on how death rates among the groups most affected by the crisis, like Black and Indigenous men, changed in 2024. A JuneJama study found that overdose deaths among Indigenous, Black, Latino, and multiracial populations were still on the rise in 2023, when they started to decline.
The Guardian contacted the CDC for comment, but they did not reply.
Although it gathers information on the drug supply, the Drug Enforcement Administration (DEA) does not disseminate specific data on things like drug purity. Only data on drug seizures that are publicly accessible at the state and federal levels are currently older than ten months. When the DEA does occasionally release reports and alerts, they frequently lag months or even years behind emerging trends in the drug supply.
The DEA reported that fentanyl purity was declining in its most recent drug threat assessment report. Kolodny pointed out that the reason for this drop in purity is probably because fentanyl is less common in Millennium Health and drug seizure data. “Fentanyl is entering the United States at a lower rate because it became more costly for the cartels to produce,” Kolodny stated. However, not every region of the nation experiences these trends. The DEA declined to comment or provide the Guardian with regionally specific fentanyl purity data.
According to researchers and public health officials who spoke to the Guardian, the DEA also doesn’t share this kind of data with them. Jenny Hua, medical director of the Chicago Department of Public Health, stated, “I don’t know why they don’t.”. I’m not sure if it’s a cultural issue. “.”.
Naturally, that data ought to be easier to access, Kolodny stated, adding that he can see why the DEA might be wary of researchers whose objectives don’t always coincide with those of law enforcement.
Although 2025 drug seizure data is not yet available, 2025 Millennium Health data indicated that fentanyl use indicators were rising in states where overdoses started to rise again this year.
When you zoom in on a local level, the picture becomes more complex, but fentanyl positivity rates in both urinalysis and drug seizure data were most strongly associated with overdose death rates.
Methamphetamine was also strongly associated with overdose death rates in a few states, including Oregon, Washington, and Ohio, the Guardian found.
When allocating funds for overdose prevention, information about the drug supply can be useful. Numerous interventions are only effective in preventing opioid-related deaths, such as naloxone. Methamphetamine is a stimulant, and the treatment for stimulant-use disorder differs greatly from that for opioid-use disorder.
Methamphetamine is the primary drug used by many Thurston County, Washington, residents, according to surveys, but finding treatment can be challenging because resources are focused on opioids, according to Katie Strozyk, who oversees the county’s opioid and overdose response program. The number of overdose deaths in Thurston County rose by 6% between the height of the crisis and December 2024.
Overdose prevention is critical.
Kolodny stated that more accurate data at the local level, much closer to real time, could be obtained if overdose prevention was given priority. There are already some US counties proving that this is feasible. Drug death toxicology reports, for instance, can be obtained in less than a month in Cook County, Illinois.
Chicago has also ramped up its tracking of the drug supply and other overdose-related information, such as ambulance data, in order to pinpoint overdose spikes at a neighborhood level and understand why they are happening.
However, the majority of the nation does not experience this kind of prompt, hyper-local response. According to Strozyk, who oversees the opioid response program in Thurston County, Washington, officials concentrate on taking a proactive stance because they “may or may never know the underlying cause” of a local spike in overdoses.
According to Hua, death rates are influenced by a variety of factors outside the purview of health departments. In a large city, where people and resources are concentrated, a coordinated response is simpler. Additionally, different regions are being affected differently by changes to the drug supply.
Additionally, a Georgia Department of Public Health spokesperson stated that they do not have local data.
Catoosa County, Georgia, had the highest rate of overdose deaths among the counties with data available for the Guardian’s analysis, with a 120 percent increase since the national peak. Meanwhile, Dougherty county, Georgia, saw one of the largest decreases, by 74 percent. The spokesperson stated that there is “no clear cause” for the discrepancy in the statewide data. Here’s an illustration of why sophisticated surveillance and data-sharing techniques are essential for giving local partners the knowledge they require to concentrate their efforts on particular counties, demographics, and overdose risk factors. “”.
Kolodny notes that methamphetamine use is prevalent in the majority of these states, making it crucial to comprehend why deaths are rising in some of them. According to Kolodny, “there are a lot of people in methamphetamine states who are addicted to both drugs [methamphetamine and fentanyl].” Perhaps the increased purity of methamphetamine is contributing to the rise in fatalities. It’s challenging to extract that. “.”.
The increase “is driven largely by synthetic opioids, especially fentanyl, which is increasingly found in combination with stimulants such as methamphetamine,” according to a Medicaid office spokesperson in Arizona, where overdose rates approached their 2025 peak. These drug combinations increase lethality and make responding to an overdose more difficult. Arizona has special challenges, such as “proximity to major trafficking routes,” according to the spokesperson. The state is using CDC funding to examine drug paraphernalia, spot new drug threats, and connect more people to substance use treatment.
According to The Guardian’s examination of data from Millennium Health, more precise patterns show up the more closely you look. Compared to state-level data, the correlation between cocaine-positive urine samples and overdose deaths is significantly higher in Cook County. In Cook County, overdose deaths are also linked to fentanyl. Hua’s group is already reacting to the city’s cocaine addiction.
“We treat these as stimulant-use disorders, whether it’s cocaine or meth,” Hua said. The health department has partnered with Northwestern University to offer “contingency management,” a highly successful treatment approach that incentivizes patients to refrain from drug use, as part of this emphasis.
Kolodny claims that the emphasis on fatalities rather than prevention is one of the reasons the situation has become so concerning.
An addiction crisis.
The US has poor drug supply and death surveillance, but even worse addiction data, Kolodny said. Prior to a few years ago, it was totally ineffective and extremely basic. You can draw conclusions from roughly 70,000 in-person interviews because it’s called the National Survey on Drug Use and Health, but that’s insufficient. “”.
An organization in Portugal is dedicated to monitoring and treating addictive behaviors. To obtain a complete picture, the agency compiles information from law enforcement, health agencies, population surveys, and even medications discovered in wastewater analysis. The rate of overdose deaths in Portugal is less than one-third that of the United States.
The agency that conducts the national survey in the United States, SAMHSA, is uncertain about its future under the Trump administration. It can be extremely challenging to work while in intensive substance-use recovery, so new regulations requiring Medicaid recipients to work 80 hours per month may also make it more difficult for patients to get treatment. Under the One Big Beautiful Bill Act, which was signed into law in July, these regulations will take effect in 2027.
Kolodny explained that deaths should be viewed as a symptom of the crisis rather than the primary result. He stated that “tens of thousands of people were dying from PCP pneumonia at the height of the Aids crisis.” “If you had Aids, it knocked out your immune system, and you got this type of pneumonia,” he said. “”.
However, Kolodny went on: “We didn’t call the Aids crisis a PCP pneumonia epidemic. We identified it as HIV, a contagious illness that can be treated with antiretroviral medication and avoided with clean syringes and condoms. But we have it backwards when it comes to the opioid crisis,” he said. If you truly understand that the majority of these deaths are in addicts rather than people who say, “Hey, shooting up fentanyl would be a fun way to spend a Friday night,” then you can see that this is an addiction crisis. It’s people who are having a lot of difficulties. “.”.
