The death of Prince was just one of more than 25,000 fatal opioid overdoses that occur yearly in the United States. Drug overdose is now the leading cause of accidental death in this country, and the majority of these events involve opioids. Most (almost 20,000 per year) are from legal opioid prescriptions.  Perversely, drugs that doctors are prescribing to help people in pain now lead to thousands of deaths. How can these drugs have become so dangerous?

Ubiquity is part of the problem. Americans represent only 7 percent of the world’s opioid consuming population, yet use around 56 percent of the world’s opioids. We far exceed the opioid consumption of any other nation, including developed countries such as the UK and Canada. With all this consumption comes tolerance to the effects of the drugs.

One might think that developing tolerance would protect a patient against dying from overdose. But opioids act through complex mechanisms, and tolerance does not develop to all drug actions at the same rate. Alcoholics can consume large amounts of alcohol yet not suffer the side effect of losing consciousness: they have become tolerant to both effects and side effects of the drug. In contrast, opioid abusers become tolerant to the analgesic (pain-killing) effects of the drug, but still suffer opioid side effects such as constipation and respiratory depression (slowing down of breathing). This is differential tolerance—tolerance develops to the main effect of the drug but not to some side-effects. Of these side-effects, respiratory depression is the most dangerous.

As anesthesiologists we see differential tolerance daily. We use powerful opioids as part of our regular toolkit to help patients wake up comfortably from surgery. One of the ways we determine how much opioid to give is to predict its effect on pain from its effect on breathing. Toward the end of an operation, when we are just about to wake a patient up, we give just enough opioid to slow the respiratory rate down to about 10 breaths per minute. In opioid-naïve patients we can then expect an adequate effect on pain as well, and in fact, we’ll typically see a comfortable wake-up and limited need for more opioid drugs in the recovery room. In these patients, opioid effect (analgesia) and side effects (respiratory depression) are fairly well matched, and we can use one to predict the other.

However, in patients who use opioids chronically this formula no longer applies. These patients have become tolerant to the analgesic effects of opioids, needing larger doses to achieve the same pain relief, but have not developed tolerance to respiratory depression. If we administer opioids at the end of these patients’ surgeries to achieve the same respiratory rate of 10, they will wake up in pain. In the worst cases, we can keep giving more and more opioid, but the patients will still rate their pain as severe – even when they stop breathing completely! It’s not uncommon to be called to "code blue" events where a patient has been found in bed not breathing, and where the problem is solved by rapid administration of naloxone (which reverses opioid effects). Often these are patients who regularly used opioid medication, and who received additional opioids for a painful procedure in the hospital. These are the victims of differential opioid tolerance: they required the additional drug to control their pain, but it then induced respiratory arrest.

This is how people such as Prince, who use opioids chronically, can overdose so easily. Once they no longer obtain pain relief, their opioid dose is increased, either by the patients themselves or by their doctor. This extra drug provides some more analgesia, but exposes the patient to the most dangerous opioid side effect—respiratory depression. In an unmonitored situation, without doctors and nurses, pain relief then can become deadly. Prince died, likely not because of too little, but because of too much tolerance to opioids.

Fixing the problem of our national opioid dependence is not easy. The Centers for Disease Control recently released a helpful guideline, acknowledging that opioids are ineffective treatment for many types of chronic pain, and asking doctors to prescribe other pain killers first, to “start low and go slow,” and to be careful in prescribing other drugs (like sedatives) at the same time, as they may compound the respiratory depression caused by opioids. In the surgical setting, we’re similarly reducing opioid use – and we’ve seen improved pain control while doing so!

Opioids are amazing drugs—in the right setting. The aftermath of major surgery would be a horror without them. Appropriate chronic use in cancer pain and end-of-life care greatly improves patients’ quality of life. For other uses, though, particularly taking opioids for chronic pain conditions not associated with cancer, we have little to suggest that they help. Instead, much evidence indicates that they harm. In those settings, we’ll have to wean ourselves, as doctors and as patients, from the habit. And as with breaking any type of dependence, this will be a long and difficult process.