On 2016 Nov 28, David Juurlink commented:

I thank Dr. Tucker for this comment. Upon re-examination, the quote above (taken from Vowles KE, 2015) does not accurately reflect the methodology of the Cochrane review. I have shared this observation with the quote's author. I remain concerned, however, about the generalizability of the review's conclusions regarding the risk of iatrogenic opioid addiction, which are at odds not only with a large body of clinical experience but also with two recent, comprehensive systematic reviews (Chou R, 2015, Vowles KE, 2015) that conclude opioid addiction occurs far more frequently than this review suggests.

On 2016 Nov 26, John Tucker commented:

There is an old saying "If it sounds too good (bad) to be true, it probably isn't". Thus my curiosity was raised when Dr. Juurlink, quoting a third party rather than the Cochrane Review itself, implied that the low rate of addiction reported in this review is an artifact of arbitrarily imputing zero addiction rate to studies in which it was not reported.

Reference to the original text shows that Dr. Juurlink's statement and the letter he quotes are both incorrect.

The actual content of the review is below:

"Six studies [of 26 total] specifically stated that no cases of addiction were observed, and 18 studies did not report whether addiction was observed.... Among the studies where addiction or addiction and abuse rates are specifically reported, the total event rate is 0.27% (7/ 2613)."

The Review further states that if the rate in the remaining trials (which were much smaller) is assumed to be zero, the addiction rate falls by about half.

Thus in contrast to Dr. Juurlink's implied statement, the conclusion of this meta analysis, that addiction rates in chronic pain are quite low among appropriately screened patients, is NOT contingent upon unrealistic assumptions.

On 2016 Nov 06, David Juurlink commented:

Regarding the risk of addiction during chronic opioid therapy, readers of this review are directed to the correspondence associated with Vowles KE, 2015, in which the authors note: "Importantly, of the 26 studies reviewed by Noble et al., only 2 (7.7%) reported rates of opioid addiction and those authors imputed (page 8) an addiction rate of zero in the other 24 studies (92.3%). Although there is clear utility in their broader findings, we would urge caution in assuming absence of any particular phenomenon simply because it is not reported."

On 2016 Nov 06, David Juurlink commented:

Regarding the risk of addiction during chronic opioid therapy, readers of this review are directed to the correspondence associated with Vowles KE, 2015, in which the authors note: "Importantly, of the 26 studies reviewed by Noble et al., only 2 (7.7%) reported rates of opioid addiction and those authors imputed (page 8) an addiction rate of zero in the other 24 studies (92.3%). Although there is clear utility in their broader findings, we would urge caution in assuming absence of any particular phenomenon simply because it is not reported."