Amoebiasis remains an important clinical problem in countries around the world, with 40 to 50 million people affected. Mortality rates are significant, with 40,000 to 110,000 deaths each year. In fact, amoebiasis mortality is second only to malaria as cause of death from protozoan parasites (Hughes 2000; Stanley 2003). Amoebiasis is prevalent in countries where public health and personal hygiene are sub-optimal (Hughes 2000; Stanley 2003). Increasing travelling, immigration of individuals from endemic areas, growth of the homosexual population, and increasing immunosuppression are factors contributing to the increased risk for amoebiasis worldwide (Hughes 2000). Endemic areas are the Indian subcontinent, Southeast Asia, Africa, and South and Central America (Reed 1992).
Ingestion of Entamoeba (E) histolytica cysts through food or water contaminated by human faeces causes amoebiasis. Asymptomatic colonisation of the gastrointestinal tract is common, but some patients develop invasive disease of the colon (Hughes 2000; Stanley 2003). The amoebae can breach the colonic mucosal barrier and travel through the portal circulation to the liver. E histolytica blocks intrahepatic portal venules, and proteolytic enzymes digest the parenchymal cells forming a liquefied central area of necrosis, this causes amoebic liver abscesses.
Amoebic liver abscess is the most common extraintestinal manifestation of E. histolytica infection. All age groups are affected, but it is 10 times more common in the 20 to 40-year old age group (Ruiz-Palacios 1997; Petri 1999) and 12 times more common in men than women. If left untreated, amoebic liver abscesses can be fatal, with death from sepsis. With early diagnosis and treatment with metronidazole alone, mortality has dropped to less than 1% (Ravdin 1995).
Ultrasound and computed tomography scans are non-invasive, equally sensitive imaging modalities for the detection of amoebic liver abscesses (Hughes 2000; Stanley 2003), but they cannot specifically differentiate amoebic from pyogenic abscess. Serum antibody detection is an important confirmatory test in the case of amoebic liver abscesses. Serologic tests are about 90% sensitive for amoebic liver abscess, with important limitations within the first week of the disease, and complementary serological assays must be performed (Petri 1999).
Metronidazole is the drug of choice for treatment of amoebic liver abscesses followed by a luminal agent to eradicate the asymptomatic carrier state (Hughes 2000). Cure rates are 95% with disappearance of fever, pain, and anorexia within 72 to 96 hours (Ravdin 1995; Reed 1998). Most abscesses heal from the periphery usually in four weeks after initiating therapy. Mean time to complete radiologic resolution is three to nine months with greater than 50% reduction in liver size within a week (Hughes 2000).
However, approximately 10% to 15% of patients remain symptomatic despite of proper drug treatment. In recent years, imaging-guided percutaneous treatment with needle aspiration or catheter drainage has replaced surgical intervention as the procedure of choice for therapeutically reducing abscess size (Tandon 1997). Simple needle aspiration is less invasive, less expensive, allows for aspiration of multiple abscesses in the same session, and avoids problems related to follow-up catheter care (Giorgio 1995; Tandon 1997). Therapeutic aspiration is the standard of care in complicated liver abscesses, which include those with high risk of abscess rupture; left lobe liver abscesses, because of increased frequency of peritoneal leak or rupture into the pericardium; no response to drug therapy within five to seven days; bacterial co-infection, and uncertainty in diagnosis (Haque 2003; Stanley 2003).
It is important in clinical practice to identify subgroups of patients with amoebic liver abscesses who will benefit from ultrasound- or computed tomography-guided therapeutic aspiration. Spontaneous rupture of the abscess into the peritoneum can occur in 2% to 7% of patients (Ravdin 1995), and mortality rates ranged from 2% to 18% (Sharma 1996). Prediction of rupture is difficult in a setting without a high index of suspicion or without ultrasound monitoring to ascertain changes in the thickness between the skin surface, the liver capsule, and the cavity margin. Existing evidence, however, on the role of image-guided percutaneous therapy is conflicting. Clinical trials are small with debatable results and we have been unable to identify systematic reviews or meta-analyses on this topic. If needle aspiration hastens response to amoebicidal drugs and clinical recovery, this can potentially improve the quality of life, shorten hospital stay, and possibly reduce health care costs especially in developing countries.
To determine the beneficial and harmful effects of image-guided percutaneous procedure plus metronidazole compared with metronidazole alone in patients with uncomplicated amoebic liver abscess.
Criteria for considering studies for this review
Types of studies
All randomised clinical trials were included in the review regardless of language, blinding, and publication status. In case there were too few randomised clinical trials, quasi-randomised clinical trials were considered for inclusion in an exploratory analysis to assess adverse events.
Types of participants
Adult patients admitted to the hospital with a clinical diagnosis of uncomplicated amoebic liver abscesses confirmed by a positive serologic result for E histolytica either by enzyme-linked immunosorbent assay (ELISA) or indirect hemagglutination assay (IHA), and with characteristic radiologic features on diagnostic imaging (computed tomography scan or ultrasound) confirming liver abscesses.
The trials that included patients with pyogenic abscesses, impending rupture, and/or other findings suggestive of complicated liver abscess, were not considered in this review.
Types of interventions
Image-guided percutaneous procedure, either needle aspiration or catheter drainage, plus metronidazole versus metronidazole in a similar dosage and duration alone.
Any collateral interventions if used equally in all intervention groups were allowed.
Types of outcome measures
- Proportion of patients without resolution of abscess size on image follow-up.
- Clinical improvement or response to treatment as follows:
Lack of resolution of fever expressed as:
- proportion of patients remaining febrile;
- days to resolution of fever;
Lack of resolution of pain expressed as:
- proportion of patients without resolution of pain;
- days to resolution of pain;
Lack of resolution of abdominal tenderness expressed as:
- proportion of patients without resolution of tenderness;
- days to resolution of tenderness;
- proportion of patients without reduction in liver size.
4. Duration of hospitalisation.
5. Adverse events. Occurrence of complications, that is, rupture of abscess cavity and number of patients requiring surgical intervention.
Search methods for identification of studies
We searched the following databases: The Cochrane Hepato-Biliary Group Controlled Trials Register (Gluud 2008), The Cochrane Central Register of Controlled Trials (CENTRAL) in The Cochrane Library (Issue 2, 2007), MEDLINE (1950 to November 2007), EMBASE (1980 to September 2007) and Science Citation Index Expanded (1945 to September 2007) (Royle 2003). The search strategies are displayed in Appendix 1. We reviewed the reference lists of the retrieved articles for potentially relevant studies, including review articles on the topic. We contacted the corresponding authors of relevant studies identified from the initial search and experts in the field for any information on unpublished articles.
Data collection and analysis
Two authors (NC, JH) independently reviewed the search output for potentially relevant trials for inclusion. Two authors (NC, JH) assessed the trials for potential inclusion. We excluded studies that do not meet the inclusion criteria and stated the reason in the 'Characteristics of excluded studies'. Disagreements were settled by discussion with a third co-author (FT). Each trial was assessed for possible multiple publications from the same data set to ensure that each trial is included only once in the review. We contacted the authors of the studies to obtain or verify missing information in the trial.
Assessment of bias risk by components of methodological quality
Two authors (NC, JH) independently assessed the bias risk by the following components of methodological quality of included studies (Schulz 1995; Moher 1998; Kjaergard 2001; Wood 2008):
Generation of the allocation sequence
- Adequate, if the allocation sequence was generated by a computer or random number table. Drawing of lots, tossing of a coin, shuffling of cards, or throwing dice were also considered as adequate if a person who was not otherwise involved in the recruitment of participants performed the procedure.
- Unclear, if the trial was described as randomised, but the method used for the allocation sequence generation was not described.
- Inadequate, if a system involving dates, names, or admittance numbers were used for the allocation of patients.
- Adequate, if the allocation of patients involved a central independent unit, on-site locked computer, identically appearing numbered drug bottles or containers prepared by an independent pharmacist or investigator, or sealed envelopes.
- Unclear, if the trial was described as randomised, but the method used to conceal the allocation was not described.
- Inadequate, if the allocation sequence was known to the investigators who assigned participants or if the study was quasi-randomised.
- Adequate, if the trial was described as double blind and the method of blinding involved identical placebo or active drugs. Due to the use of image-guided aspiration or catheter drainage as an intervention, we are well aware that it may be very difficult to properly blind such trials.
- Unclear, if the trial was described as double blind, but the method of blinding was not described.
- Not performed, if the trial was not double blind.
- Adequate, if the numbers and reasons for dropouts and withdrawals in all intervention groups were described or if it was specified that there were no dropouts or withdrawals.
- Unclear, if the report gave the impression that there had been no dropouts or withdrawals, but this was not specifically stated.
- Inadequate, if the number or reasons for dropouts and withdrawals were not described.
Furthermore, we registered whether or not the randomised clinical trials used 'intention-to-treat' analysis (Gluud 2001) and sample size calculation.
Any disagreement was resolved by discussion and settled by a third author (FT). We contacted the trial author for clarification as necessary.
Two authors (NC, JH) independently extracted data for each of the outcomes from each relevant study using a pre-designed data extraction form. We extracted the following information on study characteristics: population studied, interventions performed, and outcomes evaluated. These include trial setting, criteria for inclusion and exclusion of participants, details on interventions performed including dose of medications, duration of treatment, and co-interventions. Outcome measures to be extracted included mortality, symptom resolution, radiologic resolution of abscess cavity, and length of hospital stay. We also extracted adverse events reported such as occurrence of complications, non-responders, and those needing surgical intervention.
Data were entered in Review Manager Version 5.0 (RevMan 2008) by one author (NC) and checked by two other authors (JH, FT). Continuous outcomes were expressed as mean differences with 95% confidence intervals while dichotomous outcomes were expressed as relative risks with a 95% CI. For each outcome we extracted the number of participants assigned to each group and whenever possible extracted data to allow for an intention-to-treat analysis. If the number randomised and the numbers analysed were inconsistent, we reported this as the percentage lost to follow-up. For binary outcomes, the number of participants experiencing the event in each group was recorded. For continuous outcomes, the arithmetic means and standard deviations for each group were extracted. Any disagreement was resolved by discussion with reference to the trial report and resolution by a co-author (FT). For outcomes for which data were not reported or were reported in a different format, we contacted the authors for clarification. We checked for heterogeneity among trials by visual inspection of the forest plots and by using the chi-squared and I
Description of studies
A total of 124 studies were retrieved from the broad search strategies used. After the elimination of editorials, reviews, and repeated reports, seven trials met the inclusion criteria, and 31 studies were excluded from inclusion in this review. The 31 excluded studies with details on why they failed to meet the inclusion criteria are outlined in the table of 'Characteristics of excluded studies'. Details of the seven included trials are outlined in the table of 'Characteristics of included studies'. All seven trials were published in English. The number of participants per trial ranged from 29 to 52 participants. Unfortunately a detailed description of included participants (eg, anatomical considerations, serological status, and a precise definition of disease status) was not provided and limits the overview of the included population.
Considering the lack of a clear definition of large abscess in the included studies, the magnitude of the abscess was diverse. In the Blessman 2003 et al trial, abscess larger than 10 cm were excluded, but in all other trials no information on the size of the abscess was stated. Additionally, a significant difference on the abscess size among groups was observed in the trial by Sharma 1989 et al, and some information about the size was provided in de la Rey Nel 1989. Most of the trials involved abscess within the range of 6 cm to 8 cm.
Six trials were performed in developing countries where amoebic liver abscesses are prevalent and one trial was performed in the United States (Van Allan 1992). All trials involved a total of 310 patients diagnosed with uncomplicated amoebic liver abscess based on clinical, serologic, and radiologic criteria. The mean age ranged from 35 to 46 years, with the majority of trials involving a greater number of males than females.
Description of interventions
Five trials employed ultrasound-guided aspiration of the abscess cavities by needle aspiration (de la Rey Nel 1989; Sharma 1989; Arredondo 1993; Tandon 1997; Blessman 2003) and two trials by insertion of a temporary catheter (Widjaya 1991; Van Allan 1992). In all trials, aspiration or drainage of the abscess cavities were performed until drainage ceased. All trials but two evaluated oral metronidazole in doses ranging from 750 mg to 800 mg three times a day over ten days. de la Rey Nel 1989 et al trial administered the drug for five days and Van Allan 1992 et al trial administered the drug for a period of three weeks. Three trials administered other luminal amoebicidal drugs as co-interventions in addition to metronidazole; that is, dehydroemetine (Arredondo 1993), iodoquinol (Van Allan 1992), and chloroquine (Widjaya 1991).
All trials but Widjaya 1991 measured clinical improvement in terms of fever lysis and resolution of abdominal pain as outcomes. Analysis of fever resolution was precluded in two trials due to incomplete reporting (Arredondo 1993; Blessman 2003). Three trials reported clinical improvement as resolution of symptoms on a daily basis (de la Rey Nel 1989) and based on a graded scale measuring severity (Van Allan 1992; Tandon 1997). The latter two trials assessed symptom resolution as a reduction from baseline level. Data from Van Allan 1992; Tandon 1997 were reported as continuous data. One trial reported on fever lysis (Sharma 1989), and one reported resolution of abdominal pain as proportions of patients experiencing symptom resolution (Blessman 2003). Two trials reported the number of days to resolution of abdominal tenderness as an outcome (de la Rey Nel 1989; Tandon 1997), and one trial reported the proportion of patients demonstrating a reduction in liver size (Sharma 1989). One trial evaluated radiologic resolution of the abscess size by serial ultrasonographic monitoring and reported these outcomes as proportions of patients with radiologic resolution of abscess cavities (Sharma 1989). Three trials measured duration of hospitalisation as an outcome (Van Allan 1992; Arredondo 1993; Tandon 1997). Five of the eight trials reported adverse events (de la Rey Nel 1989; Sharma 1989; Widjaya 1991; Van Allan 1992; Arredondo 1993), but the report was incomplete, precluding analysis.
Risk of bias in included studies
The included trials varied in methodological quality, which in general was low (Figure 1; Figure 2). A description of the methodological quality of each of the included trials is given in the table of 'Characteristics of included studies'. All trials failed to report randomisation procedures in sufficient detail (and additional information was not provided). Only one trial was considered to have adequate allocation concealment (Van Allan 1992). Allocation concealment was unclear in five trials (de la Rey Nel 1989; Sharma 1989; Widjaya 1991; Tandon 1997; Blessman 2003). In one trial, the methodological quality was not possible to be clearly determined (Arredondo 1993). Blinding of outcome assessment was unclear in six trials (de la Rey Nel 1989; Sharma 1989; Widjaya 1991; Arredondo 1993; Tandon 1997; Blessman 2003) and not blinded in one trial (Van Allan 1992). All trials reported a per-protocol analysis. Four trials reported number of participants lost to follow-up (de la Rey Nel 1989; Sharma 1989; Widjaya 1991; Blessman 2003), and three trials did not specify whether there was any lost to follow-up (Van Allan 1992; Arredondo 1993; Tandon 1997). It was observed an important heterogeneity in the outcomes reported, which is an important limitation to pool the trials as their result will be highly biased.
|Figure 1. Methodological quality graph: review authors' judgements about each methodological quality item presented as percentages across all included studies.|
|Figure 2. Methodological quality summary: review authors' judgements about each methodological quality item for each included study.|
Effects of interventions
The results of the included trials are described in the table of 'Data and analyses' and are described below.
Van Allan 1992 et al reported that there were no deaths. The other six trials did not report any data on deaths.
Proportion of patients remaining febrile
Two trials reported the number of patients remaining febrile as an outcome, with an aggregate sample size of 78 patients (aspiration plus metronidazole = 37, and metronidazole = 41) (Sharma 1989; Van Allan 1992). The study population, methods of intervention, and outcomes measured in these two trials were similar enough in order to combine them in a meta-analysis. No statistical heterogeneity was found with a Chi square = 0.22, df = 1, P = 0.64, I
|Figure 3. Forest plot of comparison: 1 Needle aspiration and metronidazole versus metronidazole alone, outcome: 1.1 Proportion of patients remaining febrile.|
Days to resolution of fever
Two trials with an aggregate sample size of 70 patients (aspiration plus metronidazole = 35, metronidazole = 35) reported this outcome (Van Allan 1992; Tandon 1997). When data from these two trials were pooled, a statistically significant heterogeneity was noted (Chi square = 13.69, df = 1, P = 0.0002, I
|Figure 4. Forest plot of comparison: 1 Needle aspiration and metronidazole versus metronidazole alone, outcome: 1.2 Days to resolution of fever.|
Proportion of patients without resolution of pain
Three trials reported this outcome with an aggregate sample size of 117 patients (metronidazole = 60, aspiration = 57) (Sharma 1989; Van Allan 1992; Blessman 2003). The pooled relative risk showed a tendency towards favouring aspiration 0.66 (95% CI 0.35 to 1.26) in terms of resolution of abdominal pain, although this was not found to be statistically significant. A medium level of heterogeneity was noted with a Chi square = 4.50, df = 2, P = 0.11, I
|Figure 5. Forest plot of comparison: 1 Needle aspiration and metronidazole versus metronidazole alone, outcome: 1.3 Proportion of patients without resolution of pain.|
Days to resolution of pain
Two trials (Van Allan 1992; Tandon 1997) with an aggregate sample size of 70 patients (metronidazole = 35, aspiration = 35) reported this outcome. When data from these trials were pooled, a statistically significant heterogeneity was found (Chi square = 10.04, df = 1, P = 0.002 I
|Figure 6. Forest plot of comparison: 1 Needle aspiration and metronidazole versus metronidazole alone, outcome: 1.4 Days to resolution of pain.|
Proportion of patients without resolution of abscess size
Two trials monitored resolution of abscess size by serial ultrasonographic monitoring with an aggregate sample size of 70 participants (metronidazole = 37, aspiration = 33) (Sharma 1989; Widjaya 1991). When data for this outcome were pooled, heterogeneity was not noted (Chi square = 0.58, df = 1, P = 0.44, I
|Figure 7. Forest plot of comparison: 1 Needle aspiration and metronidazole versus metronidazole alone, outcome: 1.5 Proportion of patients without resolution of abscess size.|
Proportion of patients without resolution of abdominal tenderness
One trial reported this outcome with a relative risk of 0.54 (95% CI 0.19 to 1.56) with a sample size of 39 participants (aspiration plus metronidazole = 20, metronidazole = 19) (Blessman 2003) Figure 8.
|Figure 8. Forest plot of comparison: 1 Needle aspiration and metronidazole versus metronidazole alone, outcome: 1.6 Proportion of patients without resolution of abdominal tenderness.|
Days to resolution of abdominal tenderness
Two trials with an aggregate sample size of 81 patients (aspiration plus metronidazole = 39, metronidazole = 42) reported this outcome (de la Rey Nel 1989; Tandon 1997). When data for these two trials were pooled, statistical heterogeneity was found (Chi square = 2.69, df = 1, P = 0.10, I
|Figure 9. Forest plot of comparison: 1 Needle aspiration and metronidazole versus metronidazole alone, outcome: 1.7 Days to resolution of abdominal tenderness.|
Duration of hospitalisation
Three trials reported this outcome with an aggregate sample size of 92 patients (Van Allan 1992; Arredondo 1993; Tandon 1997). The results of Arredondo 1993 and Tandon 1997 were not consistent with the Van Allan 1992 trial, favouring metronidazole treatment alone. When data for these three trials were pooled, a medium level of heterogeneity (Chi square = 11.32, df = 2, P = 0.003, I
|Figure 10. Forest plot of comparison: 1 Needle aspiration and metronidazole versus metronidazole alone, outcome: 1.8 Duration of hospitalisation.|
Number of patients requiring surgical intervention and rupture of abscess cavity
Two trials reported adverse events as proportion of non-responders to initial intervention (de la Rey Nel 1989; Arredondo 1993). When data from these two trials were meta-analysed, a statistically significant heterogeneity was noted (Chi square = 7.84, df = 1, P = 0.02, I
|Figure 11. Forest plot of comparison: 1 Needle aspiration and metronidazole versus metronidazole alone, outcome: 1.9 Proportion of non-responders.|
The decision to therapeutically decrease abscess size in uncomplicated amoebic liver abscess is an area of controversy and we address this issue in our review. We found no evidence that image-guided needle aspiration provides added benefits in the management of uncomplicated amoebic liver abscess. However, small study populations, inadequate methods to minimize bias, and significant heterogeneity made interpretation of pooled estimates problematic. Most of the trials included in this review were performed in the late 1980s and early 1990s in low-income countries where amoebic liver abscesses are endemic. Evidence shows that the quality of randomised clinical trials affects estimates of intervention efficacy, which is significantly exaggerated in low-quality trials. Kjaergard et al reported the association between trial quality, trial setting, and year of publication (Kjaergard 1999). Most of the trials included in the present review failed to incorporate methodological procedures, which restrict bias. This lack of rigorousness may bias estimates of treatment effect. Inadequate sample size, unclear methods of generation of allocation sequence to ensure adequacy of randomisation, inability to conceal treatment allocation, and lack of blinding allow exaggeration of treatment efficacy when results of these high-risk bias trials are meta-analysed. To minimise bias, adequate methodological approaches in trial design, conduct, and reporting of results when assessing therapy of amoebic liver abscesses are needed to obtain robust conclusions. However, the clinical expertise and knowledge about the intrinsic limitations of statistic inferences must be another important tool to interpret the available information (Guller 2008).
In addition, significant heterogeneity has cast doubt on the robustness of conclusion drawn from these trials. Possible sources for heterogeneity in this review are variability in timing of outcome assessment, variability in definition of outcomes to be assessed, presence of co-interventions, and differences in the general quality of care received as a consequence of trial setting. Trials varied in defining the most relevant clinical outcome, ie, is it symptomatic improvement or is it radiologic resolution. Improvement in symptoms is no doubt important; however, these symptoms have to be clearly defined and timing of assessment must be uniformly evaluated.
Furthermore, it is equally important to study hard clinical outcomes, such as mortality and adverse events related to the treatment intervention, which can greatly influence treatment success. However, other outcomes, albeit more subjective, are more likely to be impacted upon by the experimental therapy, as observed in the analyses on resolution of pain and tenderness. Drug-related adverse events, as well as treatment-related complications as a result of invasive procedures, such as needle aspiration, were not explored in greater detail in the trials included in this review.
This lack of uniformity and inadequate methodological approach in clinical evaluation reflect the lack of standardisation in the therapeutic approach to amoebic liver abscesses. The creation of a diagnostic and therapeutic algorithm for amoebic liver abscesses may contribute to create this much-needed uniformity in therapy and the potential to give rise to well-designed clinical trials in the future.
Implications for practice
This review found no evidence to support or refute aspiration of the abscess cavity plus metronidazole versus metronidazole alone in uncomplicated amoebic liver abscess. Trials were, however, small and lacked methodological rigour for us to be able to conclude that aspiration does or does not benefit patients. Accordingly, our findings are inconclusive to make a definite recommendation on the benefit of adjunctive therapeutic needle aspiration.
Implications for research
Randomised clinical trials with larger sample sizes and adequate randomisation (generation of the allocation sequence as well as allocation concealment) and blinded outcome assessment of outcomes important to patients are urgently needed. Such trials should use uniform measures to assess outcomes.
Strict evaluation of adverse events resulting from different interventions employed in the management of amoebic liver abscesses should be included in future trials. Furthermore, trials ought to be reported according to the recommendations of the CONSORT statement.
We recognise the valuable efforts of E Labio, R Destura, MM Alejandria, and MLO Daez who started on the preparation of the review after publishing its protocol (Labio 2004), but due to other obligations could not finalise it and had abandoned it before the external peer reviewing process began. Parts of the text from the abandoned version of the review were kept by the new team of authors, though they too had to undergo modifications and updates during the process of developing the present review.
We thank Kate Whitfield for her assistance in the search strategy and retrieval of full-text journal articles, and Dimitrinka Nikolova and Christian Gluud of The Cochrane Hepato-Biliary Group for ongoing support for this review.
We thank Sandra Garcia-Osogobio (Department of Surgery, Medica Sur Clinic & Foundation) for her criticism and corrections.
We thank the CHBG Editors RL Koretz, USA; U Güller, Canada; BR Davidson, UK for the helpful comments.
Peer Reviewers: A Merens, France; NA Deepak, India.
Contact Editor: C Gluud, Denmark.
Data and analyses
- Top of page
- Authors' conclusions
- Data and analyses
- What's new
- Contributions of authors
- Declarations of interest
- Sources of support
- Differences between protocol and review
- Index terms
Appendix 1. Search strategies
Last assessed as up-to-date: 4 April 2008.
Protocol first published: Issue 3, 2004
Review first published: Issue 1, 2009
Contributions of authors
Norberto C Chavez-Tapia. Co-ordinating, data collection, designing search strategies, undertaking searches, screening search results, screening retrieved papers against eligibility criteria, extracting data from papers, writing to authors of papers for additional information, data management for the review, entering data into RevMan, analysis of data, writing the review.
Jorge Hernandez-Calleros. Data collection, undertaking searches, screening search results, screening retrieved papers against eligibility criteria, extracting data from papers.
Felix I Tellez-Avila. Data collection, designing search strategies, undertaking searches, screening search results, screening retrieved papers against eligibility criteria, extracting data from papers, analysis of data, writing the review.
Aldo Torre-Delgadillo. Interpretation of data, providing a clinical perspective, and providing general advice on the review.
Misael Uribe. Performing previous work that was the foundation of the current review, providing general advice, and securing funding for the review.
Declarations of interest
Sources of support
- None, Not specified.
- None, Not specified.
Differences between protocol and review
The protocol of this review was published with a title 'Metronidazole with or without image-guided percutaneous procedure for uncomplicated amoebic liver abscess'. We have modified it into 'Image-guided percutaneous procedure plus metronidazole versus metronidazole alone for uncomplicated amoebic liver abscess' for clearer wording.
In the excluded studies, left lobe abscesses was also considered as an exclusion criteria.
The Science Citation Index Expanded was included in the searching for identification methods.
Data analysis was performed with Review Manager Version 5.0 instead of Review Manager Version 4.2.
Additional information was requested to all authors by e-mail or conventional mail.
Medical Subject Headings (MeSH)
Antiprotozoal Agents [*therapeutic use]; Combined Modality Therapy [methods]; Liver Abscess, Amebic [*drug therapy; *surgery]; Metronidazole [*therapeutic use]; Randomized Controlled Trials as Topic; Suction [*methods]; Ultrasonography, Interventional
MeSH check words
* Indicates the major publication for the study