氟哌啶醇用于治疗姑息治疗患者的恶心呕吐
摘要
研究背景
恶心和呕吐是绝症和不治之症患者的常见症状。恶心和呕吐都会令人痛苦。氟哌啶醇通常用于缓解这些症状。这是最初的Cochrane系统综述的更新版本，该综述发表于2009年第2期，关于氟哌啶醇用于治疗姑息治疗患者的恶心和呕吐。
研究目的
评估氟哌啶醇治疗姑息治疗患者恶心呕吐的疗效和不良事件。
检索策略
对于此更新的系统综述，我们在2013年11月和2014年11月对CENTRAL，EMBASE和MEDLINE进行了更新检索。我们在2015年3月检索了对照试验注册库，以识别任何正在进行或未发表的试验。我们没有进行语言限制。对于原始系统综述，我们使用相关的检索词和同义词在2007年8月进行了数据库检索，包括CENTRAL，MEDLINE，EMBASE，CINAHL和AMED。手工检索对原始系统综述的电子检索进行了补充（使用纳入研究的参考文献列表，相关章节和综述文章）。
纳入排除标准
我们考虑纳入氟哌啶醇在任何情况下用于治疗恶心或呕吐或两者兼有的随机对照试验（randomised controlled trials, RCTs）。研究对象必须是接受姑息治疗或患有不治之症的成年人。我们排除了认为恶心或呕吐或两者均源于妊娠或手术的研究。
资料收集与分析
我们从每个电子数据库中将记录导入书目包中，并将它们合并到一个核心数据库中，在其中筛查标题、关键字和摘要的相关性。如果无法确定通过摘要确定是否纳入或拒绝，我们将获得本文的全文以供进一步评估。两位系统综述作者独立评估研究是否符合纳入标准。作者之间在研究评估方面没有意见分歧。
主要结果
我们纳入了2007年检索到的27项研究。在此更新中，我们纳入了2013年检索到的38项研究和2014年检索到的两项研究。我们确定了一项符合本次更新的纳入标准的中等质量、低偏倚风险的RCT，在22名受试者中比较了用于腕部的ABH (Ativan®，Benadryl®，Haldol®）凝胶与安慰剂缓解恶心的效果。ABH凝胶包括氟哌啶醇以及苯海拉明和劳拉西泮。在这个小型研究中，凝胶没有显著优于安慰剂；然而，据报道氟哌啶醇局部应用时吸收不明显，因此该试验没有解决氟哌啶醇通过其他途径（如口服、皮下或静脉）给药时是否有效或耐受性良好的问题。我们确定了一项正在进行中的氟哌啶醇用于治疗癌症患者恶心和呕吐的试验，其初步结果发表在会议摘要上表明氟哌啶醇对65％的患者有效。在我们进行系统综述时时，该试验结果尚未完全发表。一项进一步的试验已经开始，比较口服氟哌啶醇和口服甲氨苄嗪（左美丙嗪）治疗与治疗无关的癌症和恶心的疗效，我们打算在下一次的综述更新中纳入这项研究。
作者结论
自本综述的前一个版本，我们发现一个新的研究并纳入，但结论保持不变。从已发表的随机对照试验中尚无完整的证据来确定氟哌啶醇在姑息治疗中对恶心和呕吐的有效性。除了ABH凝胶剂与安慰剂的试验以外，我们没有发现任何完全公开的RCTs探索氟哌啶醇对姑息治疗患者的恶心和呕吐的有效性，但发现了两项正在进行的试验。
PICOs
简语概要
氟哌啶醇用于治疗姑息治疗患者的恶心呕吐
氟哌啶醇通常用于帮助控制恶心（感觉病态）或呕吐（病态），这两种疾病都是严重威胁生命的疾病的患者的常见问题。氟哌啶醇可以通过口服或注射给药。有一些研究正在探索这种药物如何在因手术引起的恶心和呕吐中起作用，以及尝试预防由抗癌治疗引起的恶心和呕吐的时机。
这是对2009年发表的原始系统综述的更新，该综述没有任何研究符合纳入标准。在本次更新中，我们在2014年11月对公开发表的文献进行了检索，发现了一项中等质量的随机对照试验，该试验将含有氟哌啶醇和其他两种药物的ABH（Ativan®，Benadryl®，Haldol®）凝胶与安慰剂进行了比较。
该试验显示ABH凝胶剂与安慰剂之间无差异。但是，先前已证明，在使用ABH凝胶后，氟哌啶醇不会被吸收，因此该结果不足为奇。我们确定了一项将氟哌啶醇用于恶心和呕吐的癌症患者的试验，该试验的初步结果在一个会议上发表。这表明氟哌啶醇对65％的患者有效，但在我们进行系统评价时，结果尚未完全发表。澳大利亚已经开展了一项进一步的试验，将氟哌啶醇与另一种用于恶心的药物甲氧并丁嗪（左美丙嗪）进行了比较。
Authors' conclusions
Background
This is an updated version of the original Cochrane review published in Issue 2, 2009, of Haloperidol for the treatment of nausea and vomiting in palliative care patients.
All patients with terminal illness should have access to palliative care, independent of their diagnosis, and we wanted to reflect this in our review. Defining this population has been identified as a problem in previous reviews. We used the same definition as Hirst 2001 to give some consistency with other Cochrane reviews: "adult patients in any setting, receiving palliative care or suffering an incurable progressive medical condition".
Description of the condition
Nausea and vomiting are common symptoms for patients with terminal, incurable illnesses. Both symptoms can be distressing. Between 6% and 68% of patients with advanced cancer are troubled by nausea ( Solano 2006 ), and nausea and vomiting are also common in other conditions, for example long‐term lung conditions and heart failure ( Edmonds 2001 ; Klinkenberg 2004 ; Solano 2006 ). There are many potential causes in patients with terminal illness, including biochemical abnormalities (for example, kidney failure, high calcium salts in the blood), drugs (for example, iron supplements or morphine), or the underlying illness (for example, cancer deposits in the liver or brain). Anxiety can also be associated with nausea. Medications used to improve nausea and vomiting are called antiemetics. Antiemetics can help control symptoms while actions are taken to try and treat the underlying cause ( Twycross 1998 ; Bentley 2001 ; Mannix 2004 ; Reuben 1986 ).
Description of the intervention
Haloperidol is in the butyrophenone class of drugs and acts as an antagonist on dopamine receptors. Haloperidol is used alone or in combination with other antiemetics orally, subcutaneously, intravenously or intramuscularly and can also be used intranasally ( Miller 2008 ). It is also available in a compound gel including lorazepam and diphenhydramine, although it is reported not to be absorbed by this route ( Smith 2012a ).
Possible side effects of haloperidol include sedation, movement disturbance and arrhythmias ( Twycross 2014 ). Neuroleptic malignant syndrome is a more serious but less common adverse event. This has numerous features including fever, altered consciousness and muscle rigidity, and can be fatal ( Susman 2001 ).
How the intervention might work
Dopamine is an important neurotransmitter in the vomiting centre in the brain. Haloperidol acts as an antagonist at dopamine receptors in the brain ( Smith 2012b ; Twycross 2014 ).
Why it is important to do this review
There is little evidence from published randomised trials for many of the drugs used for these symptoms in this patient group (for example, cyclizine, haloperidol or levomepromazine) ( Glare 2004 ). Haloperidol is commonly used in this setting to treat nausea and vomiting ( Critchley 2001 ; Smith 2012b ; To 2014 ). A small uncontrolled study (42 participants with cancer and nausea or vomiting unrelated to cancer treatment) suggests some evidence for effectiveness in the setting: 61% of evaluable participants had partial or complete control of nausea at day two (47% of all participants) and 74% at day five (40% of all participants) ( Hardy 2010 ). We felt it important to update the previous systematic review to establish the current evidence base from randomised trials ( Perkins 2009 ).
Objectives
To evaluate the efficacy and adverse events associated with the use of haloperidol for the treatment of nausea and vomiting in palliative care patients.
Methods
Criteria for considering studies for this review
Types of studies
Randomised controlled trials (RCTs) of haloperidol for the treatment of nausea or vomiting, or both, in any setting.
Types of participants
Inclusion criteria
Adults receiving palliative care or suffering from an incurable progressive medical condition. Adults suffering from nausea or vomiting, or both.
Exclusion criteria
Nausea or vomiting, or both, thought to be secondary to pregnancy or surgery.
Types of interventions
Studies where haloperidol was used as an antiemetic (alone or in addition to other agents) including any dose of haloperidol, via any route, over any duration of follow‐up.
Acceptable comparators
-
Placebo.
-
Other drug.
-
Non‐pharmacological intervention.
Types of outcome measures
Primary outcomes
-
Patient‐reported nausea severity.
-
Patient‐reported vomiting severity.
As there is a wide variety of instruments to measure these symptoms, we accepted any measure.
Secondary outcomes
-
Quality of life measurement.
-
Acceptability of treatment.
-
Need for rescue antiemetic medication.
-
Adverse events.
-
Withdrawal from study because of side effects.
Ideally, valid outcome measures would have been used but we would not have excluded studies on the basis of their outcome measures.
Search methods for identification of studies
For the original review we searched the electronic databases including CENTRAL, MEDLINE (1950 to August 2007), EMBASE (1980 to August 2007), CINAHL (1981 to August 2007) and AMED (1985 to August 2007), using relevant search terms and synonyms. The basic search strategy was ("haloperidol" OR "butyrophenone") AND ("nausea" OR "vomiting"), modified for each database. For the original review, handsearching complemented the electronic searches (using reference lists of included studies, relevant chapters and review articles). We did not impose a language restriction on studies. See Appendix 1 for the MEDLINE search strategy. We performed database searching in August 2007. For this update, we performed updated electronic searches of CENTRAL, MEDLINE and EMBASE in September 2012 and again in November 2013 and in November 2014.
We also searched clinical trials registers, the WHO International Clinical Trials Registry Platform ( http://www.who.int/ictrp/en/ ), ClinicalTrials.gov , Current Controlled Trials ( www.controlled-trials.com/ ) and the EU Clinical Trials Register ( clinicaltrialsregister.eu ) for this update on 06 March 2015, using the search term "haloperidol".
Electronic searches
We searched the following databases without language restrictions.
-
The Cochrane Central Register of Controlled Trials (CENTRAL) (via The Cochrane Library , 2007) for the previous version, and (Issue 10 of 12, 2014) for this update.
-
MEDLINE (via Ovid) 1946 to August 2007 for the previous version, and November 2014 for this update.
-
EMBASE (via Ovid) 1974 to August 2007 for the previous version, and November 2014 for this update.
-
CINAHL (via EBSCO) 1981 to August 2007 for the previous version.
-
AMED (via Ovid) 1985 to August 2007 for the previous version.
We used medical subject headings (MeSH) or equivalent and text word terms. There were no language or date restrictions. The search strategies for CENTRAL, MEDLINE, EMBASE, CINAHL and AMED are in Appendix 1 .
Searching other resources
We searched the metaRegister of controlled trials (mRCT) ( www.controlled-trials.com/mrct ), clinicaltrials.gov ( www.clinicaltrials.gov ) and the World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) ( http://apps.who.int/trialsearch/ ) on 06 March 2015 to identify additional completed or ongoing studies.
Data collection and analysis
For the original review, records were imported from each of the above electronic databases into the bibliographic package EndNote 9 and merged into a core database where titles, keywords and abstracts were inspected for relevance. If it was not possible to accept or reject an abstract with certainty, the full text of the article was obtained for further evaluation. We reviewed the updated searches from 2013 and 2014 in a Word document including title, keywords and abstracts.
Selection of studies
Two review authors independently assessed abstracts and possible studies for inclusion in accordance with the above inclusion criteria (PP and SD for the review published in 2009 ( Perkins 2009 ) and FM‐B and SD for this updated review). There were no differences in opinion between review authors with regards to assessment of studies.
Data extraction and management
Data were entered into RevMan 5.3 ( RevMan 2014 ).
We planned to assess studies for treatment effect (see Types of outcome measures above), specifying numbers needed to treat for an additional beneficial outcome (NNTB) and numbers needed to treat for an additional harmful outcome (NNTH).
Assessment of risk of bias in included studies
We graded studies that met the inclusion criteria in the Risk of bias table below (see Characteristics of included studies ).
Measures of treatment effect
Treatment effect was given as the mean difference in the change in nausea scores from baseline ( Fletcher 2014 ). Other estimations of treatment effect would also have been considered.
Unit of analysis issues
We accepted randomisation of the individual patient.
Dealing with missing data
Had missing data been potentially relevant to the findings of the review, we would have contacted the authors to include the missing data if available; if unavailable, we would have used imputation (e.g. last outcome value carried forward).
Assessment of heterogeneity
Palliative care populations can be diverse, including a range of diagnoses as well as demographic characteristics. We did not exclude studies on the basis of scales used, raising the possibility of several different outcome measures being used. If the heterogeneity of studies allowed we would have performed a meta‐analysis. If the I 2 statistic value was > 50% we would have used a random‐effects model ( Higgins 2003 ).
Assessment of reporting biases
We planned to assess heterogeneity using L’Abbé plots ( L’Abbé 1987 ), a visual method for assessing differences in results of individual studies.
Data synthesis
Quantiative meta‐analysis would only be used if studies were sufficiently similar to do so on the basis of I 2 statistical test to assess heterogeneity ( Higgins 2003 ). In the qualitative synthesis, risk of bias is noted.
Subgroup analysis and investigation of heterogeneity
If there had been sufficient data we had planned to perform subgroup analyses using the following subgroups identified a priori:
Population subgroups
-
Diagnoses of patients.
-
Likely mechanism of nausea/vomiting.
-
Prognoses of patients.
-
Age of patients.
-
Gender of patients.
Intervention subgroups
-
Route of administration.
-
Drug dose (< 2 mg/24 hours; 2 mg to 5 mg/24 hours; > 5 mg/24 hours),
Outcome subgroups
-
Nausea.
-
Vomiting.
Sensitivity analysis
Sensitivity analysis was not undertaken as there were insufficient studies identified to warrant it.
Results
Description of studies
Results of the search
We obtained 27 full studies (one in German) as potentially fitting the inclusion criteria for the 2009 review; we obtained a further 38 studies in 2013 as potentially fitting the inclusion criteria. None of these studies met the criteria for inclusion (see Figure 1 ). A further two studies were obtained from the search in November 2014 including one randomised study ( Fletcher 2014 ) and one published as a conference abstract ( ACTRN12610000481077 ).
Study flow diagram.
We also identified the two ongoing trials and one uncompleted trial below in trials registers, from 190 records (163 trials) in the WHO International Clinical Trials Registry Platform, 137 records in ClinicalTrials.gov, 18 records in Current Controlled Trials and 33 records in the EU Clinical Trials Register (date of searches 06 March 2015):
Trials in progress
-
A two‐stage trial of response to antiemetic therapy in patients with cancer and nausea not related to anticancer therapy. Study 1 is a randomised, open‐label study of guideline‐driven, targeted antiemetic therapy versus single‐agent antiemetic therapy ( ACTRN12610000481077 ). Haloperidol is the treatment in arm 2 of this study (escalated in a three‐step schedule from 1 mg/24 hours to 3 mg/24hours orally or subcutaneously). It is also used in combination with dexamethasone for a subset of arm 1 (participants with mechanical bowel obstruction). Preliminary results have been published as a conference abstract ( ACTRN12610000481077 ). The authors conclude that haloperidol is effective for 65% of patients, with no significant difference between haloperidol or targeted antiemetic therapy, however the full results were not published at the time of our review.
-
A study for participants with cancer who experience ongoing nausea, not related to their treatment, despite taking standard and usual medications, that studies the effectiveness of oral methotrimeprazine (levomepromazine) versus oral haloperidol. This study was registered 23 February 2015, with a target sample size of 126. Patients will be randomised to receive blinded encapsulated oral methotrimeprazine (6.25mg) or oral haloperidol (1.5mg) given once daily for three days, with a potential to increase to twice daily if there is no response at 24 hours or 48 hours. Responses will be assessed using a 0 to 10 numeric rating scale of nausea ( ACTRN12615000177550 ).
Trial stopped early
-
A study of olanzapine versus haloperidol for the relief of nausea and vomiting in patients with advanced cancer (registered 7 June 2005 but stopped early due to poor recruitment on 30 June 2008; Pereira 2012 ).
Included studies
Fletcher 2014 compared topical "ABH Gel" (Ativan® (lorazepam), Benadryl® (diphenhydramine) and Haldol® (haloperidol)) with placebo gel in 22 patients with nausea. Participants had cancer, or had a consultation with the palliative care team; patients receiving chemotherapy were excluded. The study sample size was reported to be large enough to detect a significant change in nausea (two points on a ten point scale), with a power of 80%. Patients were randomised to a sequence of treatments: one group that applied the ABH gel initially and then the placebo; the other group applied the placebo first followed by ABH gel. Twenty patients completed both treatments. It is unclear which arm was completed by the two participants who dropped out, ABH gel or placebo. The study is of moderate quality overall (GRADE criteria, Schünemann 2011 ).
Excluded studies
In total 27 studies were excluded in the initial review ( Perkins 2009 ) with a further 37 studies excluded in this update ( Excluded studies , with reasons shown below in Characteristics of excluded studies ).
Risk of bias in included studies
The comparison of ABH gel with placebo ( Fletcher 2014 ) recruited 22 participants of whom 20 completed the trial. The study sample size was calculated using a paired t‐test to be adequate to show that placebo was not inferior to ABH gel (power 80%, common standard deviation 1.5 at a one‐sided significance level of 5%). Risk of bias is summarised in Figure 2 and Figure 3 .
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Allocation
In the study of ABH gel ( Fletcher 2014 ), participants were allocated a sequence of treatments randomly, using a randomisation list generated by the study statistician.
Blinding
The study of ABH gel ( Fletcher 2014 ) is described as double‐blind. Participants received ABH gel or placebo gel in randomised order. Participants and investigators were not aware of which treatment was being used.
Incomplete outcome data
Two of the 22 participants did not complete the study ( Fletcher 2014 ). It is not clear which arm of the study they were in. This is unlikely to affect the overall results of the study.
Selective reporting
Placebo is reported to be non‐inferior to ABH gel ( Fletcher 2014 ). Reporting bias appears unlikely in this instance.
Other potential sources of bias
There is some risk of recruitment bias in Fletcher 2014 : all participants were reported to have an active cancer diagnosis, and some were recruited from a bone marrow transplant clinic. The study population may not therefore fully reflect the heterogeneity of patients seen by palliative care services. Also, some potential participants with more severe nausea may have chosen not to take part in the study because they preferred a subcutaneous or intravenous injection of medication for nausea, with a likely more rapid onset of action.
Effects of interventions
Haloperidol in the form of ABH topical gel did not reduce nausea or vomiting any more than placebo.
The primary study outcome was nausea which was self‐assessed using a 0–10 scale, with zero being no nausea and 10 being the worst possible nausea. The number of episodes of vomiting over time was recorded and participants completed the Memorial Symptom Assessment Scale‐Condensed which was used to determine the secondary outcomes and side effects. Participants were also asked at the end of each treatment period, "Did you feel the treatment was effective?,” and “Did you have any side effects from the drug?” On completion of both arms of the study, participants were asked, “Which drug helped you more?,” “Which drug did you think was the real ABH gel?,” and “Which was the placebo?” ( Fletcher 2014 )
Primary review outcomes
Patient‐reported nausea: no significant difference was found between placebo and ABH gel rubbed into the wrists, with both treatments showing a small reduction in nausea scores at 60 minutes (1.7+/‐ 2.05 for ABH gel, 0.9 +/‐ 2.45 for placebo; not statistically significant P value = 0.42).
Patient‐reported vomiting: observed episodes of vomiting were recorded rather than patient‐reported. ABH gel did not reduce vomiting more than placebo (not statistically significant P value = 0.34), but most patients did not have episodes of vomiting in any case.
Secondary review outcomes
Quality of life: not measured (note short term study).
Acceptability of treatment: of the 21 participants answering the question, seven thought the ABH gel was effective and 14 thought it was not. One reported a side effect of drowsiness.
Need for rescue antiemetic medication: not reported.
Adverse events: of 21 participants answering the question "Did you have any side effects from the drug?", 20 said they had no side effects from the drug and one said they did (drowsiness). One participant did not answer the question.
Withdrawal from study because of side effects: none (two participants did not complete both arms of the study, said to be because they did not want to wait to complete the study; they denied side effects) ( Fletcher 2014 ).
Discussion
This is an updated version of the original Cochrane review published in Issue 2, 2009, on Haloperidol for the treatment of nausea and vomiting in palliative care patients.
For this update, we identified one small RCT of moderate quality which compared ABH gel, containing haloperidol, diphenhydramine and lorazepam, with placebo ( Fletcher 2014 ). However haloperidol is not absorbed significantly from application of ABH gel ( Smith 2012a ) so this study does not add to the evidence base for the effectiveness of haloperidol as an antiemetic. Other than this ( Fletcher 2014 ) there are no fully published randomised controlled trials of haloperidol for nausea or vomiting in a palliative care population, although it is frequently used by palliative care physicians ( Prommer 2012 ; Smith 2012b ; To 2014 ).
A previous systematic review of haloperidol in this context retrieved case reports and case series only ( Critchley 2001 ). In a systematic review of antiemetics in the treatment of nausea in far‐advanced cancer only case series were found to support the use of haloperidol ( Glare 2004 ). A review of the management of nausea and vomiting in people with cancer and other chronic diseases described haloperidol as "likely to be beneficial" based on consensus ( Keeley 2007 ). Indeed haloperidol is thought by some to be one of four essential drugs for the management of symptoms at the end of life ( Lindqvist 2013 ).
However, initial results from a trial comparing haloperidol with a targeted antiemetic approach suggests that either approach is effective for 65% of patients ( ACTRN12610000481077 ); the full results were not published at the time of our review. This is similar to the findings from a previous uncontrolled study which suggested a response rate of 61% at day two (47% on intention to treat analysis ( Hardy 2010 ). It is encouraging that despite the difficulties encountered by some researchers trying to conduct RCTs of the effectiveness of haloperidol, two randomised studies are currently underway ( ACTRN12610000481077 ; ACTRN12615000177550 ).
There are RCTs of haloperidol in post‐operative nausea and vomiting ( Barton 1975 ), gastrointestinal disorders ( Christman 1974 ; Robbins 1975 ), prophylaxis against nausea and vomiting associated with radiotherapy ( Cole 1974 ) and chemotherapy ( Neidhart 1981 ). A meta‐analysis calculated that the number needed to treat for an additional beneficial outcome (NNTB) (with 2 mg) to prevent postoperative nausea or vomiting compared with placebo was four ( Buttner 2004 ). It is not clear to what extent these studies are applicable to palliative care populations, although they may help to inform practice ( McLean 2013 ). The causes and mechanisms of nausea and vomiting may be somewhat different in palliative care. This may have an impact on the effectiveness of interventions.
Summary of main results
For this update, we identified one RCT which compared ABH gel, containing haloperidol, diphenhydramine and lorazepam, with placebo ( Fletcher 2014 ). However haloperidol is not absorbed significantly from topical application of ABH gel ( Smith 2012a ) so this study does not add to the evidence base for the effectiveness of haloperidol as an antiemetic. Other than Fletcher 2014 there are no fully published randomised controlled trials of haloperidol for nausea or vomiting in a palliative care population, although two randomised studies are currently underway ( ACTRN12610000481077 ; ACTRN12615000177550 ).
Overall completeness and applicability of evidence
There is incomplete evidence from published RCTs to determine the effectiveness of haloperidol for nausea and vomiting in palliative care. The study of ABH gel ( Fletcher 2014 ) does not refute the apparent anti‐emetic effect of haloperidol in clinical practice and studies in other settings ( Christman 1974 ; Cole 1974 ; Barton 1975 ; Robbins 1975 ; Neidhart 1981 ; Buttner 2004 ; McLean 2013 ) since haloperidol is not absorbed by this route ( Smith 2012a ).
Haloperidol remains frequently prescribed as an antiemetic by palliative care physicians ( Prommer 2012 ; Smith 2012b ; To 2014 ).
Quality of the evidence
Only one published RCT of moderate quality has been identified in this updated review ( Fletcher 2014 ). This is a small study of ABH gel (applied topically) compared with placebo. Since haloperidol is not absorbed via this route ( Smith 2012a ), this study neither supports nor refutes the role of haloperidol as an antiemetic.
Potential biases in the review process
The review methods sought to minimise bias by conducting a thorough search of the published literature to identify relevant studies, using predefined criteria to select studies for inclusion, and independent review by two authors. The review authors sometimes prescribe haloperidol as an antiemetic in palliative care settings.
Agreements and disagreements with other studies or reviews
A previous systematic review of haloperidol in this context retrieved case reports and case series only ( Critchley 2001 ). In a systematic review of antiemetics in the treatment of nausea in far‐advanced cancer only case series were found to support the use of haloperidol ( Glare 2004 ). A review of the management of nausea and vomiting in people with cancer and other chronic diseases described haloperidol as "likely to be beneficial" based on consensus ( Keeley 2007 ). Indeed haloperidol is thought by some to be one of four essential drugs for the management of symptoms at the end of life ( Lindqvist 2013 ).
Although we did not systematically review the evidence in other contexts, we found RCTs of haloperidol in post‐operative nausea and vomiting ( Barton 1975 ), gastrointestinal disorders ( Christman 1974 ; Robbins 1975 ), prophylaxis against nausea and vomiting associated with radiotherapy ( Cole 1974 ) and chemotherapy ( Neidhart 1981 ). A meta‐analysis calculated that the number needed to treat for a beneficial outcome (NNTB) (with 2 mg) to prevent postoperative nausea or vomiting compared with placebo was four ( Buttner 2004 ). It is not clear to what extent these studies are applicable to palliative care populations, although they may help to inform practice ( McLean 2013 ). The causes and mechanisms of nausea and vomiting may be somewhat different in palliative care. This may have an impact on the effectiveness of interventions.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.