Dead Wrong: What West Virginia’s OD Deaths Tell Us About U.S. Opioid Policy

The false narrative refuses to die.

For those of you who are only vaguely familiar with what has long been called the "opioid crisis," prepare to be surprised. If your information source is the news (or worse, Netflix), you might hear the terms "West Virginia," "overdoses," and "opioids" and think: Why haven't we shut down those damn pill mills already? 

It's a reasonable reaction for anyone who hasn't paid close attention, including most medical professionals. A physician friend recently asked me what I was writing about. I asked him to guess how many people in West Virginia died from an oxycodone overdose in 2023. He guessed 10,000.     I told him 'one.' He blinked. 'One thousand, right?' No — just one. As in, practically zero. He was astounded. 

Although the primary killer has changed, some in the media and "policy world" are still chasing pill mills, while fentanyl runs the show. That disconnect is only deepened by the latest misleading Netflix drama.

Here's where the myth starts to fall apart
Figure 1 summarizes in tabular and graph form, what is now a widely-known trend that applies to West Virginia and, with some variation, to the rest of the country: A modest but steady decrease in deaths involving prescription opioids between 2015 and 2023 during a time when nationally, the number of prescriptions written dropped by half and illicit fentanyl took over as the driver of overdose deaths. (Note: oxycodone and hydrocodone were chosen as comparators because they are (by far) the most widely prescribed opioid pills.)

Figure 1 (Top) WV deaths related to illicit fentanyl, oxycodone, and hydrocodone between 2015 and 23. (Bottom) During this same time, the number of opioid prescriptions nationwide was reduced by half—sources: Table created from data provided by the West Virginia Department of Health. All Graphs were generated by ChatGPT using data provided by Statista or WVDH.

Creating a large problem from a "small" one: Overdose trends by year

It may seem intuitive that reducing the availability of prescription opioids would lead to fewer overdose deaths. However, the data tell a different story. According to the "iron law of prohibition," something Dr. Jeffrey Singer and I have written about, "the harder the enforcement, the harder the drugs." The consequences of the iron law could not be more evident than in West Virginia (Figure 2).

Figure 2. The iron law of prohibition at work. Whatever modest benefit may have come from the crackdown on prescriptions was utterly overwhelmed by the surge in (illicit) fentanyl deaths. Although this cannot prove cause and effect, common sense tells me otherwise.

What you can and cannot conclude from West Virginia data

 WVDH breaks down overdose deaths by year, county, and the drug involved. For example, if X County reports that 100 drug overdoses involved heroin, this does not mean that heroin caused 100 deaths. It means that the drug was found to be involved with (aka related to) the deaths of 100 people who died from any drug overdose in which heroin was present. Since most overdose deaths involve multiple substances, we can only estimate that heroin was responsible for between zero and 100 deaths in that county. This trend will become important below when we discuss prescription drugs, such as hydrocodone (Vicodin) and oxycodone (Percocet, OxyContin).

Why the double counting?

It is essential to understand how these deaths are counted. According to VWDH: "Due to the fact that most drug overdose deaths involve multiple substances (polypharmacy), any individual death may involve multiple types of drugs."

Percentages exceed 100% because most deaths involve multiple drugs. Each drug detected is counted as a death, regardless of whether other drugs are present, resulting in an inflated value for each individual drug. Despite this shortcoming, this method is reasonable because it is impossible to determine which (if any) of multiple drugs caused the death. It's an imperfect system, but it provides valuable information.

Prepare for a shock: 2023 vs 2015. Progress?

A link on a West Virginia Department of Health website lets you download 25 years of drug overdose data as Excel data sheets. Each sheet is a 60 X 22 matrix, containing about 1,300 data points–a total of roughly 33,000 between 2000 and 2025. After an overdose of torturous extraction, manipulation, and graphing, the tables below were born. Even I was astounded by the "new face" of the opioid crisis.
 

What is the result of the crackdown on opioid prescriptions? Figure 3 absolutely screams the answer: 

Figure 3. Top 13 Drugs Involved in Overdose Deaths in West Virginia, in 2015 and 2023.
The figures in this spreadsheet are from an analysis completed on May 14, 2025, by the West Virginia Health Statistics Center (HSC). Source: West Virginia Department of Health.

Summary of data from Figure 3:

• The total number of deaths more than doubled.
• Drug abuse patterns changed — and not for the better
• In 2015, overdoses related to oxycodone (183) and hydrocodone (113) were significant contributors to the death count. Cocaine (100) and methamphetamine (49) were less so.
• By 2023, oxycodone (35 deaths) and hydrocodone (42 deaths) had fallen well down the list. Each drug accounts for only about 3% of overdose cases; they are no longer major contributors.
• But deaths related to cocaine (228) doubled, and those from methamphetamine (760) rose by 15-fold.
• These two alone easily eclipsed the total of all drug-related deaths in 2015
• During this time, overdoses from prescription drugs (red) declined, but these were replaced (and then some) by street drugs.
• Gabapentin was the only prescription drug to make the top five in 2023. This is mind-boggling because the drug is one of the ineffective "substitutes" for the prescription opioids that pain patients can no longer obtain.

Perhaps even more jaw-dropping is that gabapentin, which is readily prescribed in place of opioids because it is not addictive (and, perhaps simply because it is not an opioid or controlled substance), was detected in more overdose deaths (129) than both oxycodone and hydrocodone taken together.

• In 2023, buprenorphine, a drug that is highly touted to treat opioid addiction (and is also an opioid), was involved in more deaths than either oxycodone or hydrocodone.
• Celexa was detected in nearly as many deaths as hydrocodone — around 40 in each case — despite being prescribed for depression, not pain. There is genuinely something wrong when the antidepressants (wrongly) prescribed to help people in pain (instead of opioids) are present more often than the drugs they are replacing.
• How anyone can claim progress for this is beyond belief.
• Yet they do.

"We still have a very long way to go. But the steep decline in overdose death is certainly very promising, and it’s something we need better answers for.”

Andrew Kolodny (presumably with a straight face), WGBH, May 29, 2025 [2]

Let's take it a step further

Let's go back to the second paragraph. My physician friend guessed 10,000 deaths, and I responded, "One." Where does this number come from? Although it's not listed in any chart or table, it can be reasonably assumed. Here's why.

According to a 2020 review in the Journal of Pain Research, only 3.4% of overdoses involving oxycodone were due to oxycodone alone. In other words, 96.6% of the time, other drugs like cocaine, alcohol, and others are present.

Let's make an assumption that your average root canal patient is going to take their Vicodin (should they be fortunate enough to get it) as directed and avoid other drugs that may be dangerous when combined with opioids. Let's also assume that people who follow doctors' orders will be in that 3.4% group. This means that of the 42 people who died from a drug overdose, about 1.5 of them died despite using only hydrocodone. That's the "one" I mentioned at the beginning. Let's round that down to zero. Same for oxycodone.

In other words, statistically, fewer than two deaths can be attributed to hydrocodone when it is used alone. Oh, PROP... Is this a big problem? 

Time for some perspective

Let's say my assumption is correct — that virtually none of the ~40 people who died with oxycodone or hydrocodone in their systems had taken any other drug or alcohol. In other words, they followed instructions exactly and still died. That would suggest nearly zero deaths from these medications when they are used alone. 

Best, worst, and absurd case scenarios

Now, let's assume that I'm wrong and all of them disobeyed their doctor's orders and did use other drugs along with their prescription hydrocodone. In this other extreme, at most, the deaths of 35 people in West Virginia are associated with hydrocodone use. The possible range is then 1 to 35, or  0.06 to 2 per hundred thousand (a common way of stating death rates). 

What do these numbers really mean? Figure 4 puts it in perspective. And then some.

Figure 4. Death rates from rare causes in the US, unless otherwise noted. [3]

You have to look long and hard to find anything funny about either the continued overdoses in the US or (even more so) the inability of pain patients to get their lives back. Yet, it is impossible not to at least snicker when you look at Figure 4. Yes, I looked up the rates of some truly ridiculous causes of death only to find that even the most absurd of them are comparable to West Virginia's "opioid crisis." Next time Kolodny "warns" us about the dangers of prescription opioids, it's only fair that we return the favor by cautioning him not to carry a live extension cord while diving in amoeba-infested water, especially after a hike following a wild mushroom dinner.

You can't be too careful.

Conclusion

THIS is why pain patients can't get the medications they need? Instead, many are forced to turn to the street, where they face a much higher risk of dying. 

The days of pill mills are long gone. There are no longer too many opioid pills circulating throughout the US; there are too few. Yet, as we continue to hear the false story of the "deadly pill," keep in mind that these pills kill virtually no one when used properly. Nor are they a gateway drug to heroin or fentanyl.

There you have it. In our worst-hit state, where pills continue to be demonized, they are doing little or no harm. The same cannot be said for the drugs that replaced them or the unnecessary suffering of millions of pain patients, not addicts, who are being deprived of the only drugs that keep them going.

This is wrong on so many levels—"public health" at its worst. Pathetic. 

NOTES:

All data below, except where otherwise noted, comes from the West Virginia Department of Health. Don't expect to go there expecting to find graphs like these. Instead, after digging through multiple complex Excel sheets, I extracted and compiled the graphs.

[1] It can be reasonably argued that it's effectively zero — that is, prescription opioids rarely cause deaths when taken as prescribed.

[2] Kolodny, Ballentyne, and others argue that reducing the number of prescription opioids will also reduce progression to more dangerous drugs. In other words, people get addicted to pills and then progress to more dangerous opioids like heroin and fentanyl. The CDC says otherwise. [CDC 2022 Opioid Surveillance Report]. Pills are no longer a significant "gateway" to harder drugs.

[3] No, I did not double-check these numbers or cite the source. If they're off by 10X, the message is the same.