This well-conducted review concluded that there was insufficient evidence to determine the existence of opioid-induced hyperalgesia in humans, except for normal volunteers receiving opioid infusions, but these data were inconsistent. The strength, quality, and consistency of the data appear to have been poor and the authors' cautious findings appear to be reliable, but there was a risk of publication bias.

To assess whether opioids induce hyperalgesia in humans.

MEDLINE, Psychological Abstracts, Science Citation Index, and NLM PDQ databases were searched up to 2007. Search terms were reported and the key pain journals, pain meeting abstracts, previous reviews, and textbooks were searched for further references. The searches were not limited by language.

Studies that measured hyperalgesia (using an acceptable measure), in humans treated with opioids, were eligible for inclusion. Those of opioid antagonists or mu agonists were excluded. Studies had to clearly state if they included opioid addicts. Those which addressed rebound headaches, pre-surgical opioids, or where withdrawal hyperalgesia could have been an issue, were also excluded.

Included studies assessed the following ten hypotheses: Opioid addicts maintained on opioids will have decreased pain threshold or pain tolerance; opioid addicts when detoxified from opioids will increase their pain threshold or tolerance; former opioid addicts will have decreased pain threshold or tolerance; decreasing, stopping, detoxifying from an opioid or rotating to a different opioid will improve pain or allodynia; chronic pain patients placed on opioids will develop a decreased pain threshold or tolerance; chronic pain patients on opioids will have decreased pain threshold or tolerance compared with those not on opioids; opioid infusion in normal volunteers or chronic pain patients will decrease their pain threshold or tolerance; opioid infusion in normal volunteers will increase secondary hyperalgesia as measured by allodynia or hyperalgesia; perioperative opioids will increase postoperative pain or opioid requirements; and placement on opioids after surgery will lead to increased opioid intake. Further details of the studies were available in online tables.

The authors did not state how the papers were selected for review, nor how many reviewers performed the selection.

The validity of case control, cohort, and correlational studies was assessed by two independent reviewers using 14 published quality criteria (defined objective; reporting of sample characteristics; participation rate at least 80% at baseline; cases and controls from the same population and cases and controls clearly defined; participation rates for cases and controls at least 80%; standardised and acceptable methods used to collect data for pain and hyperalgesia; exposure measured similarly for cases and controls; homogeneous population and unbiased allocation methods; smallest group with over 50 participants; acceptable control; and prospective study). Each study was awarded a percentage quality score which was calculated by counting the number of positive criteria, dividing by 14 (the total number of criteria) minus the number of criteria that were not applicable, and multiplying by 100. Studies scoring less than 65% were not used to derive conclusions. Discrepancies were resolved by consensus.

One reviewer extracted the results of the hyperalgesia assessment and a second checked this information for accuracy. Discrepancies were resolved through consensus.

Studies were grouped according to which of the 10 pre-specified hypotheses they addressed, then combined in a narrative synthesis accompanied by summary tables of data. For each hypothesis a summary finding was derived, along with the percentage of studies supporting the finding, the type and consistency of the evidence (according to the criteria of the Agency for Health Care Policy and Research), and the average quality score of the studies.

In total 48 reports were included. These included 30 studies and 18 case reports or case series studies, some of which assessed more than one hypothesis. Of the 30 studies, two had quality scores below 65%.

Consistent evidence: The strongest evidence of hyperalgesia (A; consistent) came from opioid infusion studies in normal volunteers; five studies (60 patients) all showed an increase in secondary hyperalgesia.

Inconsistent evidence: There was inconsistent evidence (C) from five studies (60% type 2 evidence and 40% type 3 evidence), with 67 patients, on a decreased pain threshold with opioid infusions in normal volunteers and chronic pain patients; one of three studies reported a decreased pain threshold and three of four studies reported a decreased pain tolerance. Three of four studies (75% type 2 and 25% type 3 evidence), with 229 patients, showed an increase in postoperative opioid requirements or pain if perioperative opioids were used.

Little or no evidence: Evidence that stopping, detoxifying from an opioid or rotating to a different opioid will improve pain or allodynia comprised 19 reports including one prospective study and 18 case reports with a total of 123 patients). All 21 cases reporting allodynia reported improvements. All 114 cases reporting pain reported improvements. Evidence was rated D (little or no evidence to support the hypothesis).

For the other hypotheses, there were too few studies to draw a conclusion or the evidence came from case reports or was uninterpretable.

There was insufficient evidence to determine the existence of opioid-induced hyperalgesia in humans, except for normal volunteers receiving opioid infusions, but these data were inconsistent.

This review assessed a number of clearly defined research questions using a broad set of inclusion criteria, particularly for the eligible study designs. A number of databases and other sources were searched for relevant data, without language restrictions, but the risk of publication bias remains unclear as the authors did not report whether unpublished studies were eligible for inclusion. Attempts were made to reduce the risk of reviewer error and bias when extracting the data and assessing validity, but the authors did not report whether similar precautions were taken when selecting the studies for inclusion. Given the variation in study design, patient populations, interventions, and outcomes, the authors' choice of a narrative synthesis appears to have been appropriate. The validity of the studies was assessed using published criteria, and further details of the individual criteria and the studies were reported in online supplements. The consistency, strength and type of evidence for each review question were assessed according to well-known criteria and found to be generally poor.

Overall, the strength, quality, and consistency of the data appears to have been poor and therefore the authors' cautious findings appear to be reliable, but there was also a risk of publication bias.

Practice: The authors did not state any implications for practice.

Research: The authors stated that further prospective controlled clinical studies to assess tolerance pre- and post-opioid placement in patients with chronic persistent pain (CPP) were required.

Not stated.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.