Systematic Review With Meta-analysis

Efficacy and Safety of Treatments for Opioid-induced Constipation

Priya Vijayvargiya; Michael Camilleri; Pooja Vijayvargiya; Patricia Erwin; M. Hassan Murad


Aliment Pharmacol Ther. 2020;52(1):37-53. 

In This Article


This systematic review was performed according to the Cochrane Handbook for Systematic Reviews of Interventions,[41] and reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. We followed an a priori established protocol.

Literature Search

A database search was completed in July 2019. An expert librarian (PJE) conducted the electronic literature search with input from the investigators and utilised Ovid MEDLINE (1946 to present) including Epub Ahead of Print, In-process and Other Non-Indexed Citations, EMBASE (1988 to July 2019) and EBM Reviews Cochrane Central Register of Controlled Trials. The detailed search strategies are available in the supplement. Key words in the search strategy included: constipation, chronic constipation, slow transit constipation AND opiate alkaloids or analgesics, opioid, opiate, AND narcotic antagonists, naltrexone, naloxone, mu opioid receptors, serotonin 5-HT4, serotonin 5-HT4 receptor agonists, guanylate cyclase coupled receptors, chloride channels OR methylnaltrexone, alvimopan, 5-HT4, prucalopride, lubiprostone, linaclotide. No limitation of language or year of publication was placed on the literature search. Further studies were added after review of the reference lists in the included studies, manual searching of PubMed and discussion with a content expert (MC). Abstracts and full-text reviews were evaluated independently by two investigators (PV and PV) in order to evaluate for inclusion and exclusion criteria. Disagreements were resolved by a third reviewer (MC).

Inclusion Criteria

We included studies where patient and investigators were blinded and randomised controlled trials that evaluated either bowel function or rates of adverse events. Bowel function outcomes were collected as continuous variable (the Bowel Function Index, number of spontaneous bowel movements, number of complete spontaneous bowel movements and specific endpoints such as stool consistency based on the 7-point Bristol Stool Form Scale), or FDA-defined responder analysis of spontaneous bowel movements and complete spontaneous bowel movements (combination of >3 spontaneous bowel movements/week or complete spontaneous bowel movements/week plus 1 spontaneous bowel movements/week or complete spontaneous bowel movements/week over baseline, respectively).

Dichotomous adverse events of interest were treatment emergent adverse events, serious adverse events, major adverse cardiac event, diarrhoea, abdominal pain and opioid withdrawal symptoms. Treatment emergent adverse event is defined as an adverse event that occurred once the medication was started and was either not present prior to medical treatment or an already present event that worsened in intensity or frequency following treatment.

Serious adverse events are defined by the FDA as adverse events that result in death, a life-threatening adverse event, inpatient hospitalisation or prolongation of existing hospitalisation, a persistent or significant incapacity or substantial disruption of the ability to conduct normal life functions, or a congenital anomaly/birth defect. Important medical events that may not result in death, be life-threatening or require hospitalisation may be considered serious when, based on appropriate medical judgment, they may jeopardise the patient or subject and may require medical or surgical intervention to prevent one of the outcomes listed in this definition.

Exclusion Criteria

We excluded studies that included healthy volunteers. We excluded duplicate publications of the same trial.

Data Extraction

Data extraction was completed independently by two reviewers (PV and PV).

The main bowel function outcomes were change in spontaneous bowel movements, complete spontaneous bowel movements, Bristol Stool Form Scale, Bowel Function Index and the FDA endpoints pertaining to spontaneous bowel movements and complete spontaneous bowel movements. Change in spontaneous bowel movements, complete spontaneous bowel movements, the Bristol Stool Form Scale and the Bowel Function Index were all continuous measures. We also documented per cent of patients who met FDA endpoint for spontaneous bowel movements and complete spontaneous bowel movements (dichotomous assessment of efficacy [responder analysis]) at the end of the trial and rates of adverse events. In addition to endpoints, we included trial medication, dose, frequency, route of administration and duration.

Assessment of Risk of Bias and Certainty in the Evidence

Two reviewers (PV and PV) independently determined the risk of bias based on the Cochrane Collaboration's tool for assessing risk of bias in randomised clinical trials.[41] Disagreements were resolved with discussion. Kappa statistic to determine agreement between the two reviewers was recorded at the initial completion of the risk of bias assessment. During the analysis, if there were any studies with high risk of bias, then a separate analysis was completed to exclude those with high risk of bias from the analysis. Publication bias was assessed with a qualitative assessment of excluded studies and those with incomplete information. (PV and MC) Statistical analysis to ascertain publication bias was not possible due to the small number of trials included per analysis.

Certainty in the evidence was evaluated using the GRADE approach.[42] Randomised trials are rated as high certainty evidence. The rating could be reduced due to methodological limitations (risk of bias), imprecision (wide confidence intervals and small sample size), inconsistency (high heterogeneity), indirectness (poor external validity) or publication bias.


Due to heterogeneity of the populations and settings of the included studies, we used the DerSimonian-Laird random-effects model as applied in the software package, Open Meta.[43]

We analysed the highest dose of the medication studied. For naloxegol, two doses (12.5 and 25 mg) are effective, approved by FDA and are used clinically. Therefore, we analysed the effects of the two doses of naloxegol separately. To prevent duplication of the placebo group, we divided the number of people in the placebo group by half.[42]

The continuous variables were analysed using difference in mean and 95% confidence interval. If the confidence interval was not available, this was calculated with the P value, according to the method described by Altman and Brand.[44] The summary odds ratio was significant if the confidence interval did not cross 0.

Dichotomous variables were analysed based on number of events in treatment and placebo arm. The odds ratio was considered significant if the confidence interval did not cross 1. We analysed each medication separately for all bowel function endpoints and adverse events, but did provide a summary analysis combining all medications for each endpoint.

We conducted subgroup analysis with meta-regression for the 12.5 and 25 mg doses of naloxegol to determine if there was a difference in efficacy.

We assessed heterogeneity between the studies using the I 2 statistic. I 2 25%, 50% and 75% represented low, moderate and high heterogeneity, respectively, that are not attributable to chance, but reflect true differences across the studies' populations or settings. P values for chi-squared test were reported in the forest plots.