Methadone Overdose Deaths: First Two Weeks



Methadone is a tricky drug to start, due to the narrow margin between therapeutic dose and fatal dose. Making it more difficult, people vary a great deal in the rate at which they metabolize methadone.  Some people have a methadone half -life as short as 15 hours, while others have half- lives as long as 60hours. The average is 22 hours. So even for people with a high tolerance to other opioids, increasing methadone too quickly can be deadly.

Methadone’s long half-life makes it good for a maintenance medication, since after stabilization, there’s not much fluctuation in the blood levels. However, the long half-life makes it more difficult to adjust the dose. The change I make in a patient’s dose today may not be fully experienced by the patient for five or more days.

The tolerance to the anti-pain effect of methadone builds faster than the tolerance to respiratory suppression, adding to the danger. When methadone is used inappropriately, patients may take more methadone to relieve pain, but by the time the pain is gone, they could easily have taken a methadone overdose.

All of this explains why the first two weeks of methadone maintenance treatment are the most dangerous. According to some studies, death rates for patients starting methadone at opioid treatment programs are actually higher during the first two weeks than when using illicit opioids. (1, 2)

Even so, it’s a risk worth taking, given the proven life-saving benefits of methadone (and buprenorphine) maintenance

Patient overdose during the first two weeks is a serious concern for doctors working at opioid treatment programs. We must do all we can to keep patients safe. It’s a fine line; if we start at too low of a dose or go up too slowly, we risk having our patients drop out of treatment. And if we increase the dose too quickly, it increases the risk of overdose…

The American Society of Addiction Medicine (ASAM) recently updated their methadone induction guidelines. In past years, doctors working at opioid treatment programs (OTPs) tended to start patients at 30-40mg and increase the dose rather quickly. Now, the expert ASAM panel recommends a starting dose of 10-30mg. If that dose isn’t sufficient to suppress withdrawal, a second dose can be given after three hours, so long as the total dose is not greater than 40mg. The expert panel recommends increasing the dose no more quickly than every five days, and no more than five milligrams at a time.

Some patients are more susceptible to overdose, and physicians should consider lower methadone starting doses for these people:

-Age over 60

-Using sedating drugs like benzodiazepines

-Regularly consume alcohol

-Are on prescription medications which can interact with methadone

-Medically fragile patients, for example patients with coronary artery disease, morbid obesity, -chronic obstructive pulmonary disease (COPD), or sleep apnea

-Have risk factors for prolonged QT interval, such as a recent heart attack, personal history of heart rhythm problems, or family history of heart disease

-Patients who have been abstinent from opioids for five or more days (e.g. recent incarceration, recent detoxification or hospitalization). These patients lose some of their tolerance and might be more prone to overdose with any opioid.


Interestingly, the degree of withdrawal that the patient has when entering treatment does not correlate with the dose of methadone they will need to get rid of withdrawal symptoms. In other words, one person in terrible withdrawal may need a smaller dose than another person with milder withdrawal. The degree of withdrawal that a patient feels is only partly due to opioid tolerance. Genetic makeup may be the reason why some people have more severe withdrawal than other people.

While I always ask my new patients how much opioid they have been using per day, that alone doesn’t determine methadone starting doses. There’s incomplete cross-tolerance between other opioids and methadone, meaning we can’t use the table of equianalgesic doses.

Last week I found an interesting article describing a large study of Canadian methadone patients, which will contribute even more to what we already know about risk during the first two weeks of methadone. This study showed which patient characteristics are associated with overdose death.

The study was done in Canada from 1994 until 2010, and covered over 43,000 patients enrolled in an opioid treatment program in those years. The study looked at all overdose deaths in this patient population and found 175 deaths deemed to be from opioids. These cases were matched with patients who entered treatment around the same time as the patient who died, creating a nested case-control study.

This study found, as expected, a higher degree of risk in the first few weeks on treatment. In this study, patients in the first two weeks of treatment were 16 times more likely to die in the first two weeks of treatment than any other time in treatment.

Psychotropic drugs were associated with a two-fold risk of overdose death overall, with antipsychotics associated with a 2.3-fold risk and benzodiazepines a 1.6-fold increased risk. Antidepressants were not associated with increased risk of overdose death. Alcohol use disorder diagnosis was also associated with a two-fold increase risk of overdose death.

Even more interesting, heart disease was associated with over five times increased risk of overdose death, and serious lung disorders (sleep apnea, COPD) were associated with a 1.7 times increase in overdose death.

This is a powerful study because it was so large.

This is information I can use. I’ve been stressing about patients whom I thought were at increased risk – those who use alcohol and benzodiazepines, and those with severe lung disease. While these patients are at higher risk, from this study it appears patients on anti-psychotics are at even higher risk. And I need to do a better job of getting patients to see primary care doctors, to screen for heart disease, which gave the highest risk of all.

As time goes on, I think we’ll get more information about which patients are at higher risk. Those patients need a higher degree of interaction with treatment center staff, and better coordination of care with mental health providers and primary care doctors. I know I plan to implement a system at the OTP where I work to make sure I see patients more often if they have the risk factors described.

Obviously any patient death is a terrible thing. Of course it’s worst for the family, but it also affects the treatment team. I feel badly for the families of those 175 patients in the Canadian study who died, but they gave us information that can hopefully help us provide better care for future patients.


  1. Caplehorn et al, “Mortality Associated with New South Wales Methadone Programs in 1994: Lives Lost and Saved,” Medical Journal of Australia, 1999 Feb 1;170(3):104-109
  2. Cousins et al, “Risks of drug-related mortality during periods of transition in methadone maintenance treatment: A cohort study,” Journal of Substance Abuse Treatment, October 2011, Vol 41(3); pp252-260.
  3. Leece et al, “Predictors of opioid-related death during methadone therapy,” Journal of Substance Abuse Treatment, Oct 2015,

14 responses to this post.

  1. Posted by Annette on August 22, 2016 at 1:58 am

    Very interesting article. Answered some of my questions. Thank you for sharing here. My daughter has been on methadone maintenance for over a year…. It has not gone smoothly into sobriety.


  2. Posted by Jo Ellen Hirsch on August 22, 2016 at 2:04 am

    Your post reminded me that even though I am treating addiction and taking a medical history I still need to consider problems that the patient may not know about. When I worked in a program for the medically indigent, which overlapped with addiction in some cases, I saw a significant amount of alcohol and cocaine related heart disease. Thanks for sharpening my awareness once again.


  3. Posted by Ronny Freedom on August 22, 2016 at 11:46 am

    Dr. Burson, can you elaborate on what you plan to do to see patients more often? How often will you see high risk patients, which risk factors will indicate having the patient be seen more often, what sort of clinical indicators will you look for when you see these patients and how will you respond to these clinical indicators?

    Thank you,


    • First of all I hope to get an EKG machine soon. Since I can read them myself, in addition to looking at the QT interval, I can look for evidence of past MI and ask these patients see a primary care doctor. Next, smokers over forty with other coronary artery risk factors probably need to be sent to primary care for consideration of a stress test. Anyone with a past history of heart disease I plan to schedule to see me every three months, and patients prescribed antipsychotics for any reason probably also need to see me every three month. These patients probably need EKGs too – many of those type drugs influence the QT interval.
      I already postpone admitting patients with uncontrolled benzodiazepine use/alcohol use until they complete a detox program, and I think this data supports that action.
      Fortunately the group I work with has some smart doctors on their advisory committee, and I plan to discuss with them too.


  4. Posted by Alan Wartenberg MD on August 22, 2016 at 3:34 pm

    There was a very good response – essentially a letter to the editor but actually a good brief article – to some of the issues raised in that consensus guideline (which I point out was based on no data but on expert opinion). Firstly, if an individual has been recently on methadone and has been off for a relatively short time (few months), and has an equivalent habit to what they had the first time around, one can increase more quickly and by higher dose levels, since their tolerance is well-established. Secondly, if patients are seen 4 hours post-dosing, and a good assessment done, people can ALSO be increased more rapidly (after 3 days instead of 5-7), and potentially by higher doses (10 mg rather than 5 mg). Safety is paramount and important, but patients becoming frustrated at still being dope-sick while paying for both the clinic and extra opioids doesn’t help, and people drop out of treatment early.

    Secondly, the Caplehorn article, while widely quoted (especially by lawyers) was from Australia, where people are dosed in pharmacies, without any nursing observation and via pharmacists who do not receive any particular training in what questions to ask or what to look for. The death rate in the first two weeks is nowhere near as high in the US with our system, particularly after rapid titrations (0 to 70 mg in a week or 10 days) have been sharply reduced.


    • Yes! Thank you. I read that article, what I would call a rebuttal. I was going to include it in my post but it was getting late and I like to post something each Sunday. So I put that off, thinking I could do another blog post with their concerns. I agree with you – in some patients, it’s safe to increase more quickly. I think we can still do that, only we need to document why we decide to do what we do, otherwise lawyers, if there’s a bad outcome, can ask us why we strayed from the ASAM guidelines.
      And I did not know that about the Caplehorn article. Good information, thank you.
      What is our rate in the U.S.? I know my state has a very high rate of “methadone overdose deaths,” but every patient on methadone who dies is proclaimed an overdose death no matter what the circumstances, so our rate can’t be valid.


  5. Posted by Alan Wartenberg MD on August 22, 2016 at 3:34 pm

    I have 2 secondlies, it should be thirdly 🙂


  6. Posted by Tracie Walker on August 22, 2016 at 7:14 pm

    Dr Burson i am greatful for your blogs. It is helpful to me and helps me to understand my loved ones own addiction


  7. I would caution anyone from reading QT intervals unless they are a experienced trained cardiologist who reads EKGs by hand (with calipers) and doing the math (to get the corrected for heart rate interval) by hand instead of just going by what the machine says the QTc is. I have seen machines be off by more than 80 milliseconds which could cause someone appropriate for life saving Methadone treatment to be turned away or have their admission postponed (while they continue to use possible fentanyl laced heroin/pills) until they get a confirmation from a cardiologist.


  8. Posted by Brian Clear MD on September 5, 2016 at 11:36 pm

    As a relatively new OTP medical director I’ve really struggled to rectify some of these inconsistencies between the major guideline documents we’re supposed to be following. The ASAM guidelines, for instance, even seem internally inconsistent when they suggest that by week 5 the patient should be in long term stabilization. By their own recommendations only the most aggressively managed patient will be above a dose of 60mg daily by week 5 which will usually not be an effective maintenance dose. The higher end of what ASAM considers a normal dose, 160mg, would take months to reach while the patient continues to experience daily withdrawal.

    Then there’s TIP 43, which in contrast supports/permits daily increases until relief at peak is achieved, and that one is still referenced by the current federal guidelines (updated in 2015, making them newer than the ASAM induction guidelines) as the go-to resource for best-practices. It seems like a medical legal perfect storm in which we’re expected to comply to multiple sets of inconsistent guidelines… yet there’s no good data I can find comparing induction protocols and how they affect outcomes (complications of induction vs treatment retention). Is there anything else out there?

    In general I also agree with Dr. Wartenberg’s thought that a compromise with Q3day increases for a standard risk patient is reasonable. Also if starting a patient below the max recommended starting dose and they still have observable withdrawal at peak on days 2-3, allowing some early daily dose titration up to 40mg as needed might be appropriate as well and would actually be a little more cautious than trying to front-load people with 40 and not allowing any titration for 5 days to stay strictly in-line with ASAM.

    Also could a link be posted to this Caplehorn article? I’m having a hard time finding it.


    • You aren’t the only one to note that at ASAM’s recommended rate of increase, it’s going to take longer to get the average patient to a stable dose.
      I’d follow the ASAM guidelines. I hate to say this, but physicians are more likely to be sued for increasing faster than recommended than slower than recommended.
      I know of two excellent physicians who were sued after induction deaths, because they started on “too high a dose” as determined by the lawyer and an expert witness. All four cases were settled in favor of the plantiff. It sucks that we even have to be concerned about this when we are trying to decide what’s best for our patients, but you and I know doctors practice defensive medicine in all specialties.
      here’s a link to the Caplehorn article:


  9. In Australia we recently had the Penington Institute Australia’s Annual Overdose Report – 2016, released. it covers some of the challenges with opioids including accidental overdose and death rates. Would like to share this infographic that summarises the findings in a visual way with you.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: