Profilaxis con antibióticos para la infección del sitio quirúrgico en pacientes sometidos a un trasplante hepático
Resumen
Antecedentes
La infección del sitio quirúrgico es más frecuente en el trasplante hepático que en otros tipos de trasplante de órganos sólidos con diferentes antibióticos. Los estudios han revelado que la tasa de infección del sitio quirúrgico varía del 8,8% al 37,5% después del trasplante hepático. Por lo tanto, la profilaxis antimicrobiana probablemente es una herramienta esencial para reducir estas infecciones. Sin embargo, la bibliografía carece de pruebas que indiquen el mejor régimen profiláctico con antibióticos que se puede administrar para el trasplante hepático.
Objetivos
Evaluar los efectos beneficiosos y perjudiciales de los regímenes con antibióticos profilácticos para la infección del sitio quirúrgico en pacientes sometidos a trasplante hepático.
Métodos de búsqueda
Se hicieron búsquedas en el registro de ensayos controlados del Grupo Cochrane Hepatobiliar (Cochrane Hepato‐Biliary Group), en el Registro Cochrane Central de Ensayos Controlados (Cochrane Central Register of Controlled Trials) (CENTRAL), MEDLINE, EMBASE, Science Citation Index Expanded y en Latin American Caribbean Health Sciences Literature (LILACS). La búsqueda más reciente se realizó el 11 de septiembre de 2015.
Criterios de selección
Todos los ensayos clínicos aleatorios elegibles que compararon cualquier régimen con antibióticos versus placebo, versus ninguna intervención o versus otro régimen con antibióticos para la infección del sitio quirúrgico en receptores de trasplante hepático, independientemente de la edad, el sexo y la razón del trasplante. Para los datos sobre los efectos perjudiciales se consideraron los estudios cuasialeatorios y otros estudios observacionales si se obtuvieron de los resultados de la búsqueda de ensayos clínicos aleatorios.
Obtención y análisis de los datos
Dos autores de la revisión seleccionaron los ensayos relevantes, evaluaron el riesgo de sesgo de los estudios y extrajeron los datos.
Resultados principales
La búsqueda electrónica identificó 786 publicaciones después de la eliminación de duplicados. A partir de esta búsqueda, solo un ensayo clínico aparentemente aleatorio, publicado en forma de resumen, cumplió los criterios de inclusión de esta revisión. Este ensayo se realizó en el Shiraz Transplant Centre, Shiraz, Irán, en el cual los investigadores asignaron al azar a 180 receptores de trasplantes hepáticos consecutivos. Se consideró que el riesgo general de sesgo del ensayo publicado en forma de resumen era alto. Los investigadores no informaron datos numéricos, aunque mencionaron que 163 participantes cumplieron los criterios de inclusión después de la asignación al azar, por lo que se incluyeron en los análisis. Lo más probable es que los 17 participantes que se excluyeron fueran receptores de trasplante hepático de alto riesgo. Los autores del ensayo concluyeron que no fue posible encontrar diferencias entre los dos regímenes con antibióticos (ceftriaxona más metronidazol versus ampicilina‐sulbactam más ceftizoxima) cuando se administraron a los receptores de trasplante hepático. Los revisores no pudieron reconfirmar los análisis debido a que, como se mencionó, los autores del ensayo no proporcionaron datos para los análisis.
Conclusiones de los autores
Aún no se conocen los efectos beneficiosos y perjudiciales de los regímenes con antibióticos profilácticos para la infección del sitio quirúrgico en el trasplante hepático. Se necesitan ensayos clínicos aleatorios adicionales bien realizados que cumplan las guías SPIRIT (Spirit Protocol Items: Recommendations for Interventional Trials) y CONSORT (Consolidated Standards of Reporting Trials) para determinar la función exacta de los regímenes con antibióticos profilácticos en pacientes sometidos a un trasplante hepático.
PICOs
Resumen en términos sencillos
Profilaxis con antibióticos para la infección del sitio quirúrgico en pacientes sometidos a un trasplante hepático
Pregunta de la revisión
Efectos beneficiosos y perjudiciales de los regímenes con antibióticos profilácticos para la infección del sitio quirúrgico en pacientes sometidos a un trasplante hepático.
Antecedentes
La infección del sitio quirúrgico es más frecuente en el trasplante hepático que en otros tipos de trasplante de órganos sólidos. Los estudios han revelado que la tasa de infección del sitio quirúrgico varía del 8,8% al 37,5% después del trasplante hepático en el que se utilizan diferentes regímenes con antibióticos profilácticos. Por lo tanto, la profilaxis antimicrobiana probablemente es una herramienta esencial para reducir la infección del sitio quirúrgico. Sin embargo, la bibliografía carece de pruebas con respecto al mejor régimen con antibióticos profilácticos que se puede administrar para el trasplante hepático.
Características de los estudios
Todos los ensayos clínicos aleatorios elegibles que compararon cualquier régimen con antibióticos versus placebo, versus ninguna intervención o versus otro régimen con antibióticos para la infección del sitio quirúrgico en pacientes sometidos a un trasplante hepático, independientemente de la edad, el sexo y el motivo del trasplante. Para los datos sobre los efectos perjudiciales se consideraron los estudios cuasialeatorios y otros estudios observacionales si se obtuvieron de los resultados de la búsqueda de ensayos clínicos aleatorios.
Las pruebas están actualizadas hasta septiembre 2015. Solamente un ensayo clínico aleatorio publicado en forma de resumen examinó los efectos de los antibióticos para la profilaxis en los receptores de un trasplante hepático.
Resultados clave
Este ensayo clínico aleatorio, que se realizó en el Shiraz Transplant Centre, Shiraz, Irán, asignó al azar a 180 participantes; de estos pacientes, 163 cumplieron los criterios de inclusión después de la asignación al azar. El ensayo no informó ningún otro dato numérico y fue considerado de alto riesgo de sesgo; la calidad de las pruebas fue baja, por lo que no fue posible tener seguridad en cuanto a las conclusiones presentadas por los autores del ensayo. Este ensayo evaluó ceftriaxona más metronidazol versus ampicilina‐sulbactam más ceftizoxima. Los autores del ensayo no han respondido a tres solicitudes enviadas por los revisores a través del correo electrónico para solicitar información detallada, aunque mencionan que no se encontraron diferencias entre los dos regímenes con antibióticos cuando se administraron para la profilaxis de la infección bacteriana en receptores de un trasplante hepático.
Calidad de la evidencia
El único ensayo identificado, que fue publicado en forma de resumen, está en alto riesgo de sesgo. Los autores del ensayo no informaron resultados como eventos adversos, tiempo de supervivencia del injerto, estancia en la unidad de cuidados intensivos, duración total de la estancia hospitalaria ni calidad de vida. Aunque los autores del estudio parecieron informar sobre algunos resultados de interés, el informe es ambiguo y no hay datos disponibles. Los autores del ensayo excluyeron de los análisis a los pacientes con alto riesgo.
Conclusiones y futuras investigaciones
Esta revisión no aporta pruebas sobre los efectos beneficiosos y perjudiciales de los antibióticos para la profilaxis de la infección del sitio quirúrgico entre los receptores de un trasplante hepático. Se necesitan ensayos clínicos aleatorios bien realizados que cumplan las guías SPIRIT y CONSORT para determinar los efectos de los antibióticos en el algoritmo para la prevención de la infección del sitio quirúrgico en pacientes sometidos a un trasplante hepático.
Authors' conclusions
Background
Description of the condition
Surgical site infection is an important cause of morbidity and mortality among hospitalised patients. Data from National Healthcare Safety Network (NHSN) hospitals show that between 2006 and 2008, a total of 15,862 surgical site infections occurred after 830,748 operative procedures ‐ an overall proportion of 2% (CDC 2011). Data from October 1997 to September 2002 from 168 English hospitals showed an overall infection proportion of 3.9 infections/100 operations (Leong 2006). Patients who develop surgical site infection generally stay much longer in intensive care units, are more likely to be readmitted to the hospital and have two times greater risk of dying than those without surgical site infection. Moreover, costs of health care are substantially increased for these patients (Bratzler 2004). Surgical site infection may be influenced by patient and operation characteristics and is defined by the Centers for Disease Control and Prevention (CDC) as superficial incisional, deep incisional and organ/space infection (Mangram 1999; CDC 2011).
Among the risk factors for surgical site infection related to patient characteristics (Mangram 1999), those who undergo solid organ transplantation have additional risks associated with immunosuppression, generally poor nutritional status, often unfavourable clinical conditions, co‐morbidities, frequent prolonged pre‐transplant hospital stay, invasive procedures and use of catheters, leading to increased colonisation with multi‐drug‐resistant organisms, presence of infections previously installed at other sites of the body and greater need for blood product transfusions (Asensio 2008; Dorschner 2014). When we consider the risk factors associated with surgical procedures (Mangram 1999), we see that patients undergoing solid organ transplantation differ by surgical complexity, leading to higher complication rates and to the possibility of graft contamination (Asensio 2008; Dorschner 2014). In addition to increased morbility and mortality, longer hospital stay and additional costs, surgical site infection in transplant patients may be associated with increased probability of graft loss (Asensio 2008; Dorschner 2014) (see Differences between protocol and review).
Surgical site infection is more frequent in liver transplantation than in other types of solid organ transplantation. Studies have shown that the rate of surgical site infection varies from 8.8% to 37.5% after liver transplantation in which different regimens of prophylactic antibiotics were used (Iinuma 2004; Asensio 2008; Garcia 2008; Hellinger 2009; Kettelhut 2010; Hellinger 2011). This fact is related to the technical complexity of the procedure, the potentially infecting milieu within the abdominal cavity and the poor medical condition of many liver recipients (Asensio 2008). Post‐transplant events such as biliary complications can occur at an early stage after transplantation and are related to the development of surgical site infection. Sundaram 2011 conducted a study on the proportion of biliary surgical complications that occur during the post‐model for end‐stage liver disease (post‐MELD) era, and reported 4.9% of patients with biliary leak, 15.4% with bile duct stricture and 9.5% with hepatic artery thrombosis. Choledochojejunal reconstruction, previous solid organ transplantation and more than four units of transfused red blood cells were found to be independently associated with the development of surgical site infection by a multiple logistic regression model in patients who developed incisional, peritoneal or intra‐abdominal abscess infection within the first month after transplantation (Asensio 2008). Bacteria commonly isolated from sites of surgical infection after liver transplantation include Acinetobacter baumannii, Enterobacter spp., Escherichia coli, Enterococcus faecalis, Enterococcus faecium, Staphylococcus aureus and Pseudomonas aeruginosa (Asensio 2008; Kusne 2008).
Description of the intervention
Antimicrobial prophylaxis is an essential tool for reducing surgical site infection. Four principles must be followed to maximise the benefits of antimicrobial prophylaxis (Mangram 1999): (1) an antimicrobial prophylactic agent should be used for all operations in which its use has been shown to reduce the number of surgical site infections as documented by adequate scientific evidence, and for those operations after which surgical site infection would represent a catastrophe; (2) an antimicrobial prophylactic agent that is safe, inexpensive and bactericidal, and that covers the most probable intraoperative contaminants for the operation should be used; (3) infusion of the initial dose of antimicrobial agent should be timed so that a bactericidal concentration of the drug is established by the time the skin is incised; and (4) therapeutic levels of the antimicrobial agent should be maintained in both serum and tissues throughout the operation, and until, at most, a few hours after the incision is closed.
How the intervention might work
The role of antimicrobial prophylaxis is to reduce the microbial contaminant burden to safe levels during the surgical procedure (Mangram 1999). Various regimens of antibacterial prophylaxis are commonly used for surgical site infection in liver transplantation, including glycopeptide plus aztreonam, glycopeptide plus quinolone, glycopeptide plus antipseudomonas penicillin, ampicillin plus third‐generation cephalosporin, amoxicillin–clavulanic acid plus quinolone, amoxicillin–clavulanic acid, cefazolin and other antibiotics. The literature lacks evidence regarding the best antibiotic prophylactic regimen that can be used for liver transplantation (Asensio 2008; Kusne 2008).
Why it is important to do this review
So far, no meta‐analysis nor systematic review has studied antibiotic prophylactic regimens as prophylaxis for surgical site infection in people undergoing liver transplantation. Benefits and harms of antibiotic prophylactic therapy need to be weighed against each other and must be established.
Objectives
To assess benefits and harms of antibiotic prophylactic regimens for surgical site infection in people undergoing liver transplantation.
Methods
Criteria for considering studies for this review
Types of studies
Randomised clinical trials, no matter the language, year of publication nor blinding.
We considered for inclusion trials published as abstracts if they could fulfil the inclusion criteria of our review. We contacted the principal authors of these trials to inquire about missing information relevant to the review.
We scrutinised quasi‐randomised studies, defined as studies using inadequate allocation assignment such as date of birth, day of the week or month of the year or the person's medical record number, or as studies allocating every alternate person (Higgins 2011c), as well as other observational studies retrieved by searches for randomised clinical trials, for their reports on harm.
Types of participants
Trials with liver transplant recipients, regardless of age, sex and aetiology.
Types of interventions
We considered trials for inclusion if they involved any antibiotic regimen versus placebo, versus no intervention or versus another antibiotic regimen for surgical site infection.
We considered antibiotics administered only by the parenteral route at any dose or duration of prophylaxis. We considered only the parenteral route because the main focus of this systematic review was to study specifically perioperative parenteral antibiotic prophylaxis. We did not intend to evaluate other prophylactic strategies for prevention or treatment of colonisation, nor the use of selective bowel decontamination. One systematic review has studied other strategies for preventing bacterial sepsis and wound complications in people undergoing liver transplantation (Gurusamy 2014). Furthermore, antifungal prophylaxis in solid organ transplantation has already been evaluated in another Cochrane systematic review (Playford 2004). Therefore, we decided that we would not study this topic.
Types of outcome measures
Primary outcomes
-
All‐cause mortality.
-
Surgical site infection, as defined by the Centers for Disease Control and Prevention (CDC) (Appendix 1).
-
Adverse events: serious adverse events. A serious adverse event, defined according to the International Conference on Harmonisation (ICH) Guidelines for Good Clinical Practice (ICH‐GCP 1997), is any untoward medical occurrence that at any dose results in death, is life‐threatening, requires inpatient hospitalisation or prolongation of existing hospitalisation or results in persistent or significant disability or incapacity, or in a congenital anomaly or birth defect. We will consider all other adverse events as non‐serious.
-
Quality of life.
Secondary outcomes
-
Time of graft survival.
-
Intensive care unit stay and total length of hospital stay.
-
Costs (as a narrative description).
Outcomes were assessed at maximal follow‐up and within 30 days after commencement of prophylaxis, according to the surgical site infection definitions provided above, if no implant (e.g. biliary prosthesis, transanastomotic catheter) was left in place, or within one year if an implant was in place. Adverse effects that occurred during use of prophylactic drugs or 30 days after onset of prophylaxis were considered, as these effects were related to their use.
Search methods for identification of studies
Electronic searches
We searched the Cochrane Hepato‐Biliary Group Controlled Trials Register (Gluud 2015), the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, EMBASE, Science Citation Index Expanded and Latin American Caribbean Health Sciences Literature (LILACS) (Royle 2003). Search strategies and time spans of searches are given in Appendix 2.
Searching other resources
We checked the reference lists of identified studies for additional citations. We contacted pharmaceutical companies, study authors and experts to request unpublished data. We looked for information about ongoing clinical trials by searching the clinical trials site of the National Institutes of Health (clinicaltrials.gov) and the trial portal of the World Health Organization (apps.who.int/trialsearch/Default.aspx).
Data collection and analysis
We followed the guidelines given in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011a) and in the Cochrane Hepato‐Biliary Group Module (Gluud 2015).
Selection of studies
Two review authors (RAMBA and CNH) independently screened studies identified by the literature search. We resolved disagreements by consulting with two other review authors (RED and ENH).
Data extraction and management
Two review authors (RAMBA and CNH) independently extracted data. We resolved discrepancies by discussion and used a standard data extraction form to extract the following information: characteristics of the study (design, methods of randomisation); participants; interventions; and outcomes (types of outcomes, adverse events). We then checked for errors before entering the data into Review Manager 5 (RevMan 2014).
Assessment of risk of bias in included studies
We defined methodological quality as confidence that the design and the report of the randomised clinical trial would restrict bias in the comparison of interventions (Moher 1998). According to empirical evidence (Schulz 1995; Moher 1998; Kjaergard 2001; Wood 2008; Lundh 2012; Savovic 2012; Savovic 2012a), risk of bias in a trial can be assessed by using risk of bias domains, described as follows.
Allocation sequence generation
-
Low risk of bias: Sequence generation was achieved using computer random number generation or a random numbers table. Drawing lots, tossing a coin, shuffling cards and throwing dice are adequate if performed by an independent adjudicator.
-
Unclear risk of bias: The trial was described as randomised, but the method of sequence generation was not specified.
-
High risk of bias: The sequence generation method was not, or could not be, random. Quasi‐randomised studies and those using dates, names or admittance numbers to allocate participants were inadequate and will be excluded for assessment of benefits but included for assessment of harms.
Allocation concealment
-
Low risk of bias: Allocation was controlled by a central and independent randomisation unit, sequentially numbered opaque and sealed envelopes or a similar method, so that intervention allocations could not have been foreseen in advance of, or during, enrolment.
-
Unclear risk of bias: The trial was described as randomised, but the method used to conceal allocation was not described, so that intervention allocations may have been foreseen in advance of, or during, enrolment.
-
High risk of bias: The allocation sequence was known to investigators who assigned participants, or the study was quasi‐randomised. Quasi‐randomised studies will be excluded for assessment of benefits but included for assessment of harms.
Blinding of participants and personnel
-
Low risk of bias: It was mentioned that both participants and personnel providing the interventions were blinded, and the method of blinding was described, so that knowledge of allocation was prevented during the trial.
-
Unclear risk of bias: It was not mentioned whether the trial was blinded, or the trial was described as blinded, but the method or extent of blinding was not described, so that knowledge of allocation was possible during the trial.
-
High risk of bias: The trial was not blinded, so that allocation was known during the trial.
Blinded outcome assessment
-
Low risk of bias: It was mentioned that both participants and outcome assessors were blinded, and this was described.
-
Unclear risk of bias: It was not mentioned whether the trial was blinded, or the extent of blinding was insufficiently described.
-
High risk of bias: No blinding and no incomplete blinding were performed.
Incomplete outcome data
-
Low risk of bias: Underlying reasons for missing data are unlikely to make treatment effects depart from plausible values, or proper methods were employed to handle missing data.
-
Unclear risk of bias: Information was insufficient to permit assessment of whether the missing data mechanism in combination with the method used to handle missing data was likely to induce bias on the estimate of effect.
-
High risk of bias: The crude estimate of effects (e.g. complete case estimate) will clearly be biased by the underlying reasons for missing data, and methods used to handle missing data were unsatisfactory.
Selective outcome reporting
-
Low risk of bias: The trial reported the following pre‐defined outcomes: all‐cause mortality, surgical site infection and adverse events. If the original trial protocol was available, outcomes should be those called for in that protocol. If the trial protocol was obtained from a trial registry (e.g. www.clinicaltrials.gov), outcomes sought should have been those enumerated in the original protocol if the trial protocol was registered before, or at the time that, the trial was begun. If the trial protocol was registered after the trial was begun, those outcomes were not considered to be reliable.
-
Unclear risk of bias: Not all pre‐defined outcomes were reported fully, or it was unclear whether data on these outcomes were recorded.
-
High risk of bias: One or more pre‐defined outcomes were not reported.
Vested interest bias
-
Low risk of bias: if trial funding did not come from parties that might have conflicting interests (e.g. an antibacterial agent manufacturer), or if any academic, professional, financial or other benefits to the person responsible for the trial were independent of the direction or statistical significance of trial results.
-
Unclear risk of bias: if the source of funding was not clear, or if it was unclear whether the person responsible for the trial stands to benefit according to the direction or statistical significance of trial results.
-
High risk of bias: if the trial's source of funding had a conflict of interest, or if any academic, professional, financial or other benefits to the person responsible for the trial are dependent on the direction or statistical significance of trial results.
If risk of bias in a trial was judged as 'low' in all domains listed above, the trial would fall into the 'low risk of bias' group. If risk of bias in assessed trials was judged as 'unclear' or 'high' in one or more of the specified domains, the trial would fall into the group with 'high risk of bias'.
As a first step, information relevant to making a judgement on a criterion was copied from the original publication into an assessment table. If additional information was provided by study authors, this was also entered into the table, and it was indicated that this was unpublished information. Two review authors (RAMBA and CNH) independently made a judgement as to whether the risk of bias for each domain was judged to be 'low', 'unclear' or 'high'. We resolved disagreements by discussion.
We recorded this information for each included trial in 'Risk of bias' tables in Review Manager 5 (RevMan 2014), and we have presented a summary ’Risk of bias’ figure and graph.
Measures of treatment effect
Binary outcomes
For dichotomous data, we planned to use the risk ratio (RR) as the effect measure with 95% confidence intervals (CIs).
Continuous outcomes
For continuous data, we planned to present results as mean differences (MDs) with 95% CIs. When pooling data across trials, we planned to estimate mean differences if outcomes were measured in the same way by investigators. We planned to use standardised mean differences (SMDs) to combine trials that measured the same outcome while using different methods.
Unit of analysis issues
All recruited participants assigned randomly to trial groups.
Dealing with missing data
An intention‐to‐treat (ITT) analysis is one in which all participants in a trial are analysed according to the intervention to which they were allocated, whether or not they received the intervention. We planned to assume that participants who dropped out were non‐responders. For each trial, we planned to report whether investigators stated if the analysis was performed according to the ITT principle. If participants were excluded after allocation, we would report any details provided.
Furthermore, we planned to perform the analysis as based on an ITT principle when possible (Newell 1992). A 'best‐worse' case scenario assumes that none of the dropouts or participants lost from the experimental arm, but all of those lost from the control group, experienced the outcome, including all randomly assigned participants in the denominator; and a 'worst‐best' case scenario assumes that all dropouts and participants lost from the experimental arm, but none from the control arm, experienced the outcome, including all randomly assigned participants in the denominator.
Assessment of heterogeneity
We planned to quantify inconsistency among pooled estimates by using the I2 statistic. This illustrates the percentage of variability in effect estimates resulting from heterogeneity rather than from sampling error (Higgins 2003; Higgins 2011b). I2 = [(Q ‐ df)/Q] × 100% test, where Q is the Chi2 statistic, and df its degrees of freedom. We planned to assess heterogeneity between trials by visually examining the forest plot to check for overlapping confidence intervals, and by using the Chi2 test for homogeneity with a 10% level of significance and the I2 statistic. We planned to classify heterogeneity by using the following I2 values.
-
0 to 40%: might not be important.
-
30% to 60%: may represent moderate heterogeneity.
-
50% to 90%: may represent substantial heterogeneity.
-
75% to 100%: shows considerable heterogeneity.
For meta‐analysis that presents significant heterogeneity greater than 75%, we planned to explore clinical and methodological causes. Only if the cause was detectable would we present the results in a subgroup or sensitivity analysis. When a meta‑analysis was not appropriate, we planned to synthesise the results in a table.
Assessment of reporting biases
Apart from assessing the risk of selective outcome reporting considered under assessment of risk of bias in the included trials, we planned to assess the likelihood of potential publication bias by using funnel plots, provided that at least 10 trials assessed similar interventions. When small studies in a meta‐analysis tend to show larger treatment effects, other causes should be considered, including bias risks, heterogeneity, artefactual causes and chance.
Data synthesis
We planned to calculate pooled estimates by using random‐effects model and fixed‐effect model meta‐analyses. We planned to present both sets of results in cases of discrepancies in the results. If no discrepancies were apparent, we would have presented the results using the random‐effects model.
Trial Sequential Analysis
Trial Sequential Analysis (TSA) (CTU 2011; Thorlund 2011) is a statistical tool that can be used to quantify the statistical reliability of data in cumulative meta‐analyses, with P values adjusted for sparse data and repetitive testing on accumulating data (Wetterslev 2008; Brok 2009; Thorlund 2009). TSA is a method that combines an information size calculation with thresholds for statistical significance and contrasts this information with the cumulative information provided by the included trials.
We planned to perform TSA on the data from trials with low risk of bias (Wetterslev 2008; Brok 2009). Outcomes planned to be analysed using TSA included all‐cause mortality and surgical site infection, no matter whether they yielded a statistically significant result in the meta‐analysis. We would have used the meta‐analytical estimate of control event proportions for trials with low risk of bias as the control event proportion in the TSA. We planned to use the intervention effect estimated in the meta‐analysis of trials with low risk of bias and to perform a sensitivity analysis by using an a priori intervention effect of 20% risk ratio reduction (Higgins 2011d).
For each TSA performed, we planned to calculate a diversity‐adjusted required information size. We planned to adjust heterogeneity by using the observed diversity factor (1/(1‐D2) and the diversity estimated (D2) among all trials with a priori assumed final heterogeneity of 50% (Wetterslev 2009). For primary outcomes only, we planned to perform separate TSAs using a risk of type I error (α) of 5% and a risk of type II error (β) of 20%, as well as analyses using a risk of type I error (α) of 1% and a risk of type II error (β) of 10%.
Subgroup analysis and investigation of heterogeneity
In the case of substantial clinical heterogeneity (I2 > 50%), we planned to use subgroup analysis to assess study results. Subgroup analyses are secondary analyses in which participants are divided into groups according to shared characteristics, and outcome analyses are conducted to determine whether any significant treatment effect related to that characteristic has occurred. When data permit, we will carry out the following subgroup analyses.
-
Comparison of trials with low risk of bias versus trials with high risk of bias.
-
Types of surgical techniques.
-
Types of donors (i.e. deceased donor or living donor transplantations).
-
Types of immunosuppression.
-
Multiple other post‐surgical complications, such as biliary leak, bile duct stricture, bowel perforation and hepatic artery thrombosis.
-
Total time of antibiotic prophylaxis.
-
Retransplantation surgery.
-
Different ages of liver transplant participants.
-
Different levels of severity according to score MELD and Child‐Turcotte‐Pugh score.
-
Prior use of antibiotics and evidence of colonisation with multi‐drug‐resistant bacteria.
-
Need for antimicrobial agents due to other procedures, infectious complications or both.
We planned to set a P value of 0.05 for the Chi2 test for subgroup differences.
Had we been able to include an adequate number of trials, we planned to conduct subgroup analyses to explore causes of heterogeneity and the robustness of review results. We also planned to perform subgroup analyses while separating trials according to numbers of withdrawals for each outcome (< 20% versus ≥ 20%).
'Summary of findings' tables
We planned to use in our review the principles of the GRADE (Grades of Recommendation, Assessment, Development and Evaluation) system (Guyatt 2008) to assess the quality of the body of evidence associated with specific outcomes (surgical site infection as defined by the Centers for Disease Control and Prevention and adverse events related to the antibiotic) and we planned to construct a 'Summary of findings' table using GRADE software (ims.cochrane.org/revman/other‐resources/gradepro). The GRADE approach is used to appraise the quality of a body of evidence according to the extent to which one can be confident that an estimate of effect or association reflects the item assessed. The quality of a body of evidence considers risk of bias within studies (methodological quality), directness of the evidence, heterogeneity of the data, precision of effect estimates and risk of publication bias.
Results
Description of studies
See Characteristics of included studies and Characteristics of excluded studies.
Results of the search
We identified 786 publications after removal of duplicates. After we initially assessed these references, we studied two in greater detail and considered them to be potentially eligible for inclusion. Following verification of these publications, we excluded one from the review (Arnow 1992) and included one in the review (Nikeghbalian 2010) (Figure 1).
Study flow diagram.
Included studies
We could identify only one trial published as an abstract. Investigators randomly assigned a total of 180 participants and included 163 participants in the analysis (Nikeghbalian 2010).
Study design
Trial authors stated in the title of the abstract that their trial was a randomised controlled trial (Nikeghbalian 2010). In the abstract text, they stated that consecutive liver transplant recipients were randomly assigned to two groups. For further text details on the trial, see the Characteristics of included studies table.
Types of interventions
The included trial evaluated ceftriaxone plus metronidazole versus ampicillin‐sulbactam plus ceftizoxime in liver transplant recipients (Nikeghbalian 2010).
Types of outcomes measured
Nikeghbalian 2010 evaluated rates of positive culture and changes in antibiotic type, fever episodes, bacterial infection at specific sites (wound, urine, respiratory tract, blood, intra‐abdominal site) during the hospital course and mortality.
Excluded studies
We excluded one study from the review (Arnow 1992). This randomised clinical trial did not fulfil the inclusion criteria, as researchers studied only the pharmacokinetics of perioperative systemic antibiotics and the microbiological effectiveness of oral nonabsorbable antibiotics and evaluated none of the proposed outcomes.
Risk of bias in included studies
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
No report described generation of randomisation sequence and allocation concealment. We sent three emails to the study authors but have not yet received a reply (Nikeghbalian 2010). We judged the trial to have unclear risk of bias.
Blinding
We judged blinding to have high risk of bias, as it seems that the attending physician and participants and assessors were not blinded.
Incomplete outcome data
Investigators excluded 17 participants after randomisation. This is why we judged this domain as having high risk of bias (Nikeghbalian 2010).
Selective reporting
It was not possible to verify selective reporting because we had inadequate information; therefore, we classified this trial as having high risk of bias (Nikeghbalian 2010).
Other potential sources of bias
Participants excluded from the trial analysis presented with high risk of surgical site infection and included those with blood requirement over four bags during the operation, individuals over 55 years or younger than two years of age, those with established diabetes mellitus and participants who expired as the result of primary non‐function or surgical complications; therefore, we classified this domain as having high risk of bias (Nikeghbalian 2010).
Effects of interventions
As we included only one trial, it was not possible to perform a meta‐analysis (Nikeghbalian 2010). However, trial authors reported no differences between ceftriaxone plus metronidazole and ampicillin‐sulbactam plus ceftizoxime for prophylaxis of bacterial infection among liver transplant recipients in terms of rates of positive culture and changes in antibiotic type, fever episodes, and bacterial infection at specific sites (wound, urine, respiratory tract, blood, intra‐abdominal site) during the hospital course and mortality.
Nikeghbalian 2010 reported on mortality and surgical site infection (wound and abdominal) outcomes but otherwise did not specify the criteria used. Trial authors also analysed the remaining above listed outcomes. However, they did not report on adverse events, quality of life, time of graft survival, length of intensive care unit stay and total time of hospital stay and costs.
'Summary of findings' table
We deemed it not justifiable to conduct GRADE assessments and construct a 'Summary of findings' table because of the abstract format of the single included trial, which lacked numerical data. Following several attempts to contact trial authors, we elected to present a narrative summary of the trial.
Discussion
Summary of main results
We conducted this review to examine the benefits and harms of antibiotic prophylactic regimens for surgical site infection in people undergoing liver transplantation. We included only one trial, which was available as an abstract and examined a total of 180 randomly assigned participants. Of these, 163 met the inclusion criteria. Overall, we assessed the included trial as having high risk of bias. This trial did not report on adverse events, time of graft survival, intensive care unit stay, total time of hospital stay nor quality of life. Study authors provided no numerical data. but mentioned that participants at high risk (i.e. those with blood requirement over four bags during the operation, individuals over 55 years or younger than two years of age, those with established diabetes mellitus and participants who expired as the result of primary non‐function or surgical complications) were excluded from the trial analysis. However, we may not be sure whether these are the 17 participants who did not meet the inclusion criteria after randomisation. Trial authors did not respond to our three email requests for trial data.
Overall completeness and applicability of evidence
Because of our comprehensive search strategy and contact with experts in the field, we are confident that we have mapped existing knowledge on comparisons of different antibiotic prophylactic regimens for surgical site infection in people undergoing liver transplantation. The applicability of this review is very weak, as the only included abstract assessed effects of ceftriaxone and metronidazole versus those of ampicillin‐sulbactam and ceftizoxime at a single centre in Iran.
Quality of the evidence
The methodological quality of the included trial published in the form of an abstract is low (Nikeghbalian 2010). Overall sample size was very small, and most risk of bias domains assessed were classified as high risk. This reflects on the conclusions drawn by the authors of this review.
Potential biases in the review process
We developed a comprehensive search strategy, handsearched the reference lists of identified studies for additional citations and contacted experts in the field. Therefore, we are confident that we have identified the only available study on this topic. However, potential biases in this review would include lack of statistical power generated by lack of data and by the abstract format of the published trial, which also caused selective reporting bias.
Agreements and disagreements with other studies or reviews
No systematic review has compared the results of antibiotic prophylactic regimens for surgical site infection in people undergoing liver transplantation. Few randomised clinical trials have investigated the benefits or harms of antibiotic prophylactic regimens for surgical site infection in these patients. Prospective studies have analysed different regimens of prophylactic antibiotics against surgical site infection in patients undergoing liver transplantation (Asensio 2008; Garcia 2008).
Asensio 2008 evaluated more than eight different regimens (glycopeptide plus aztreonam, quinolones plus glycopeptide, quinolones plus amoxicillin‐clavulanate, ampicillin plus third‐generation cephalosporin, amoxicillin‐clavulanate, glycopeptide plus antipseudomonal penicillin, cefazolin and others); the most common were amoxicillin‐clavulanate and combinations of glycopeptide and antipseudomonal penicillin and glycopeptide and aztreonam. Risk of surgical site infection associated with different antibiotics ranged from 1.7% for combination glycopeptide plus aztreonam to 17.1% for cefazolin alone. Investigators found that use of cefazolin was associated with risk of surgical site infection in the univariate analysis. However, this association could not be proven when the treatment effect was controlled by centre and by Child‐Pugh class. Study authors considered it inadvisable to use cefazolin as a single prophylactic antibiotic. Furthermore, Asensio 2008 suggested that amoxicillin‐clavulanate or the combination of a third‐generation cephalosporin plus amoxicillin could serve as a reasonable antibiotic prophylactic regimen in liver transplantation, as the institutions of researchers presented a low incidence of penicillin‐resistant and vancomycin‐resistant enterococci. Asensio 2008 also considered vancomycin an antibiotic prophylactic regimen in hospitals with a high incidence of methicillin‐resistant Staphylococcus aureus post‐operative infection. This statement corroborates the results of the non‐randomised clinical study of Calleja 1993, showing that vancomycin appeared to be an elective prophylactic antibiotic during an epidemic outbreak of methicillin‐resistant Staphylococcus aureus (MRSA) infections affecting individuals undergoing liver transplantation.
Garcia 2008 compared cefazolin and amoxicillin‐clavulanate as prophylactic regimens for surgical site infection in 167 consecutive adult patients (age > 18 years) who underwent orthotopic liver transplantation. Three participants underwent combined liver and kidney transplantation. A total of 56 episodes of surgical site infection occurred. Ninety‐four recipients received cefazolin, and 73 amoxicillin plus clavulanate, as antimicrobial prophylaxis, with rates of 0.36 and 0.3 episodes of surgical site infection per transplantation, respectively (P value > 0.05). Univariate and multi‐variate analysis showed that antibiotic therapy given during the three months before orthotopic liver transplantation was the unique variable associated with surgical site infection. Study authors concluded that the incidence of morbidity caused by surgical site infection continued to be elevated, and permanent vigilance on epidemiology was the only way to choose an appropriate prophylactic protocol and empirical treatment because of incessant outbreaks of new aetiological agents. They recommended an integral vision and a multi‐factorial approach other than antimicrobial prophylaxis for prevention of surgical site infection. Although they noted no relationship between surgical site infection and mortality, investigators in Garcia 2008 found that individuals with surgical site infection had longer hospital and intensive care unit stays, and that given the cost per day of prolonged hospitalisation, prevention of surgical site infection would mean a reduction of US$ 4320 per non‐infected patient.
The incidence of multi‐drug‐resistant (MDR) and extensively drug‐resistant (XDR) bacteria has recently increased among patients undergoing solid organ transplantation. Methicillin‐resistant Staphylococcus aureus (MRSA) infection is declining, but vancomycin‐resistant enterococci, MDR/XDR Enterobacteriaceae and MDR/XDR non‐fermenters are progressively growing among patients undergoing solid organ transplantation and represent a global threat (Cervera 2014). These infections, mainly with the 'ESKAPE' pathogens (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa and Enterobacter spp.), worsen patient and graft survival and increase the costs associated with solid organ transplantation. Their treatment and control have important limitations associated with pharmacokinetics, effectiveness, recurrence, adverse effects and costs (Sifri 2012; Cervera 2014). When patients who are bacteraemic as the result of MDR pathogens receive inappropriate or late treatment, mortality is increased (Cervera 2014). Unfortunately, the literature provides little information about antibiotic prophylaxis for patients undergoing liver transplantation who are colonised with MDR or XDR bacteria, or who will receive an organ from a donor suspected or confirmed to be infected or colonised with such bacteria. Most available observational studies and specialist opinions suggest that, when possible, donors with positive blood culture with MDR or XDR, as with other bacteria, should be adequately treated before the time of donation, and recipients should be submitted to the same treatment over 14 days on the basis of in vitro susceptibility testing (Ariza‐Heredia 2012; Bishara 2012). Papers highlight the importance of inter‐institutional communication and co‐operation in providing enough time to guide adequate antibacterial treatment, even when standard prophylaxis has been initiated (Ariza‐Heredia 2012; Sifri 2012). Articles have described case reports with good outcomes and some with unfavourable outcomes (Ariza‐Heredia 2012; Herati 2012; Sifri 2012). So, whether organs should be accepted from donors with elevated risk factors for MDR or XDR bacteria, or from patients colonised or infected by these micro‐organisms, remains an unanswered question. Additional studies are needed to better resolve these matters.
The great number of antibiotic prophylactic regimens used in these studies provides evidence of lack of agreement in this area. Different regimens of prophylactic antibiotics, lack of numerical data in the abstract and exclusion of patients at high risk from the only included trial (Nikeghbalian 2010) make comparison with other studies impracticable.
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.
