Quantity of Research Available
The literature search identified a total of 367 citations. Of these, 352 citations were excluded during the title and abstract screening and 15 potentially relevant articles were retrieved for full-text review. In addition, 3 potentially relevant reports were retrieved by a literature search from other sources (i.e. grey literature). Five studies were included in the review including two SRs, one RCT, and two non-randomized studies.
A PRISMA diagram demonstrating the study selection process is presented in Appendix 1.
Additional references of potential interest are provided in Appendix 2.
Summary of Study Characteristics
Two SRs,6,7 one RCT,4 and two non-randomized studies3,5 were included in this review. Detailed study characteristics are provided in Appendix 3.
The SRs originated in Canada6 and the United Kingdom.7 Both aimed to examine the management of constipation in patients prescribed opioids for either chronic-non cancer pain or cancer pain, with Ahmedzai and Boland7 observing the effectiveness and harms of many laxative regimens (including docusate) and Ruston et al.6 primarily focusing on the efficacy and harms of docusate sodium, sennosides, and lactulose compared to PEG.7 The SR by Ahmedzai and Boland,7 which included prospective and retrospective RCTs, SRs, and comparative cohort studies, identified one SR that met their inclusion criteria. This SR included four RCTs on docusate calcium compared to placebo; however, further data was not reported from three of the four RCTs due to weak methodology (reasons included: no definition of constipation or evaluation of constipation prior to start of study, no statement on how randomization was performed, and no analysis by intention-to-treat).7 Therefore, they included data from only one RCT (N=22).10 The other SR by Ruston et al. did not identify any relevant evidence.6
Two of the primary studies examined the management of constipation3,4 and the other observed both the prevention and management of constipation.5 Two of the three primary research studies originated in Canada4,5 while the other originated in Norway.3 Two primary studies observed elderly populations in either a hospice4 or nursing home 3 setting while the other examined patients that had been admitted to a cancer center for treatment.5 Patients in the hospice setting were close to the end of life (mean age of approximately 72 to 75 years),4 those in the nursing home had various chronic co-morbidities (mean age of 85.6 years),3 and those in the Hawley and Byeon study had many different types of cancer (mean age of approximately 59 to 63 years).5 In addition, all of these studies had a certain proportion of their populations receiving opioids for either chronic non-cancer pain3,4 or cancer pain.3–5 The RCT by Tarumi et al.4 observed patients receiving oral docusate sodium with sennosides versus sennosides alone for the management of constipation (N=74) while the sequential cohort study by Hawley and Byeon5 examined either orally administered sennosides alone or sennosides plus docusate sodium treatment regimen for the prevention and management of constipation (N=60). Fosnes et al.3 observed a cross-section of elderly patients admitted to a nursing home treated with one of a number of different laxatives administered regularly or on demand (N=197). These laxatives were categorized according to the Anatomical Therapeutic Chemical Classification System (ATC) level five and included softeners/emollients (liquid paraffin), contact laxatives (bisacodyl, senna glycosides, and sodium picosulphate), bulk laxatives (ispaghula), osmotic laxatives (lactulose and macrogol combinations), and enemas (docusate sodium and laurilsulfate suppositories were included in this category).
Summary of Critical Appraisal
Details of the critical appraisal are provided in Appendix 4.
Both SRs6,7 reported rigorous methodology that included descriptions of comprehensive literature searches, with Ruston et al.6 additionally performing prescoping searches and Ahmedzai and Boland7 obtaining numerous alerts regarding harms data. In addition, both had clearly defined a priori research questions and inclusion criteria. Ruston et al.6 also clearly described the data selection and extraction methods using a two-author system and the statistical analysis that would have been used if any evidence had fit their inclusion criteria. The SR by Ahmedzai and Boland7 included grading of their evidence; an interesting caveat to this being that the grading was performed on the outcomes and populations of interest only and not necessarily on the methodological quality of the included studies themselves. The one study that provided information on the effectiveness of docusate for their SR was an RCT.7 Authors from both reviews declared their conflicts of interest.6,7
All three of the primary studies had clearly defined objectives, outcomes of interest, and reported baseline patient characteristics.3–5 Standardized or specified tools that recorded stool frequency,4 consistency,3,4 and volume4 were used in two of the three studies3,4 while the third study used a chart review.5 The Tarumi et al.4 study included a sample size calculation in order to ascertain appropriate statistical power;4 however, an important limitation to note included the lack of a specified allocation concealment in this RCT.4 Two non-randomized studies had inadequate sample sizes3,5 while all three involved numerous healthcare providers which could have potentially introduced confounding into the administration of the laxatives or the accurate reporting of symptoms or outcomes.3–5 Docusate use was not observed in many patients in the Fosnes et al. study3 and, therefore, not much could be determined regarding its effectiveness. Hawley and Byeon reported discrepancies between their two cohorts whereby the second cohort had a larger population of those with one specific type of cancer thus potentially confounding the results.3 Authors in two of the studies declared conflicts of interest.3,4 Conflict interests were not declared in the Hawley and Byeon study,5 however, specifics in funding were declared.
Summary of Findings
Detailed findings are provided in Appendix 5.
Four of the five citations identified for this review did not report any increased clinical effectiveness at either reducing symptoms associated with or for the prevention of constipation upon the administration of docusate sodium3–5 or docusate calcium.7 The fifth review was unable to identify RCTs that examined the use of docusate sodium, sennosides, or lactulose compared to PEG for the management of opioid-induced constipation.6
The included publications focused on the management of constipation in populations (or subsets of the populations) that were either being treated with opioids for malignancy pain3–5,7 or for pain associated with other chronic non-malignant disease.3,4,7 In addition, the management of constipation in some patients not taking opioids was included in the analysis of two primary studies.3,4 In patients prescribed opioids, docusate (calcium7 or sodium4) appeared no more effective than placebo7 or sennosides alone (when observing a sennosides plus docusate sodium versus sennosides alone protocol4) at increasing stool frequency, softening stool consistency,4,7 in lessening perceptions of difficulty associated with evacuation,4 in perceptions of evacuation completeness,4 or for relieving others characteristics associated with opioid-induced bowel dysfunction (e.g. abdominal cramps or delayed gastric emptying).7 In one of the primary studies involving a stepwise (increased dosing) schedule involving either sennosides and docusate sodium or sennosides alone in cancer patients, the frequency of bowel movements increased in those taking the sennosides alone protocol5. In addition, a sub-analysis involving symptom control / supportive care showed that patients following the sennosides alone protocol had statistically significantly more bowel movements and had bowel movements on more than 50% of days when compared to those following the docusate and sennosides protocol (62.5% versus 31.6%, respectively).5 Normalization of stool frequency or consistency was not achieved in 41% of nursing home residents that were treated with one of numerous laxatives regimens in the cross sectional study; however, this examined many laxatives and the results for those treated with docusate were not specified.3
Bowel care interventions, whereby rescue medications were provided to induce bowel movements if the laxative regimens were not successful, were either reported in the included studies4,5,10 or not specified (due to the cross sectional study design which observed numerous laxative regimens).3 The Tarumi et al. study4 did not observe any statistical significance in these reported interventions when comparing the docusate plus sennosides to the sennosides alone groups in hospice patients (68.6% versus 74.4%, respectively). In contrast to this, Hawley and Byeon5 reported a statistically significant 57% of cancer patients in the docusate plus sennosides protocol requiring additional interventions (lactulose, suppositories, or enemas) when compared to cancer patients following the sennosides alone protocol (40%) in their symptom control/supportive care subanalysis.5
Limitations
While the number of reviews included in the Ahmedzai and Boland7 SR was large, only one RCT (that was part of the SR they included in their analysis) informed their conclusions regarding the use of docusate for the prevention or management of constipation. This RCT10 was determined to be of lower quality and included only 22 patients, many of which received additional interventions apart from their laxative protocol.
Sample sizes were also an issue in two of the three primary studies included in this review.3,5 In the study by Fosnes et al.3 there was a large cohort of patients included in the study; however, only five patients were actively using docusate to manage their constipation. Another aspect similar to other studies included in this review, was the fact that patients were able to access additional bowel interventions if necessary.4,5,10 While these interventions were primarily provided to those whose treatment regimens appeared insufficient, this still may have introduced confounding as, perhaps, the treatments had not yet had enough time to work or the effect was due to the rescue medications and not the docusate.
As previously stated, Fosnes et al.3 observed numerous laxative regimens and did not focus on one specific laxative; therefore, little can be inferred with regard to the clinical effectiveness of docusate from this study. Information from this cross sectional observation appeared to be representative of the nursing home population (e.g. various underlying medical conditions, elderly bedridden and ambulatory) yet its incorporation of the frail and mentally reduced most likely reduced the data quality (as some patients themselves provided information on their bowel movements).3 Furthermore, this study primarily observed laxative “types” (i.e. osmotic, stimulant, etc.) and categorized them according the ATC level five criteria. Therefore, according to these criteria, docusate (administered as a suppository) was classified in the enema category3 and thus this may reduce its generalizability in a population taking oral docusate.
The RCT by Tarumi et al.4 compared the clinical effectiveness of docusate and sennosides to sennosides and placebo. While there was no clear benefit of the addition of docusate to sennosides for the management of constipation, one must take into account the potential issues associated with this hospice population. These patients were primarily at the end of their life (as the RCT reported that the median length of stay was 16 days and most were discharged at death4) and their bodies may have been in the process of shutting down. Numerous items to consider would have been whether the patients were eating and, if they were, what the consistency of their food was like. These aspects, along with disease-stage, could all have played a role in the state of constipation or in the clinical effectiveness of the aforementioned treatments. The authors did note that the lack of statistical significance between the groups may not definitively indicate that docusate was no better than placebo for the treatment of constipation as there was an inability to control for additional interventions, other medication use, and there was no internal measure of the assay sensitivity.4
Another additional issue relating to all three of the primary studies3–5 was that observations of the treatments were done at numerous settings. This introduces variation in how nurses and researchers attain information (even in the presence of standardized or validated tools). Hawley and Byeon5 specifically mentioned the diversity in the healthcare providers and how, being an un-funded study, the investigators could not be continually present to ensure consistency in reporting. In addition, even though all staff were alerted to the ongoing study, many healthcare providers would have been unaware that their patients were involved; thus increasing the possibility that the reporting of bowel movements may not have been accurately ascertained.5 In addition, two of the primary studies reported the use of standardized or specified tools used to analyze various aspects of the stool frequency, consistency, and volume.3,4 However, none of these types of tools were specified in the Hawley and Byeon study; alternately, they assessed outcome data by nursing chart review.5 Therefore, it is uncertain whether inconsistencies may have been introduced by reporting this way.
