Scolaris Content Display Scolaris Content Display

Base de datos Cochrane de Revisiones Sistemáticas Protocolo - Intervención

Pharmacological interventions for treating dyslipidemia in patients with HIV infection

Declaraciones de intereses de los autores

Versión publicada: 08 diciembre 2010 Historial de versiones

https://doi.org/10.1002/14651858.CD008869

Esta versión no es la más reciente

ver la versión actual 02 febrero 2018
Contraer todo Desplegar todo

Abstract

This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:

To assess the clinical effectiveness and safety of statins, ezetimibe, fibrates, or fish oil for treating dyslipidaemia in HIV‐infected patients receiving highly active antiretroviral therapy. Clinical effectiveness will be measured in terms of prevention (primary and secondary) of cardiovascular events (Fatal or non‐fatal myocardial infarction, stroke and angina).

Background

With the advent of highly active antiretroviral therapy (HAART) (Appendix 1), the clinical course of Human Immunodeficiency Virus (HIV) infection has been transformed from an almost uniformly fatal disease to a chronic and manageable condition, with an impressive  extension of life expectancy. However, the benefits of HAART  are tempered by a broad variety of side effects, including a wide range of laboratory and clinical disturbances.

The risk of cardiovascular disease (CVD) among HIV infected people  increased compared with that in uninfected persons (Lundgren 2008a).  In HIV‐infected patients, contributors to CVD may include traditional cardiovascular risk factors (e.g., age, sex, diabetes, hypertension, cigarette smoking, and hypercholesterolaemia) (Lifson 2010), and direct or indirect effects of HIV infection itself (including inflammation and immune activation), or adverse effects of HIV therapy (David 2002; Sudano 2002). Moreover, the type of antiretroviral therapy may also contribute through its impact on metabolic and body fat parameters. Low CD4+T cell count has been suggested as a risk factor for CVD in the HIV outpatient study (Lichtenstein 2010).

HAART‐associated metabolic syndrome (Appendix 1) is an increasingly recognized clinical entity (Canh 2010; Pullinger 2010; Adeyemi 2008; Mondy 2007; Tomazic 2004). The National Cholesterol Education Program’s Adult Treatment Panel III report (ATP III) identified the metabolic syndrome as a multiplex risk factor for cardiovascular disease (CVD) (NCPE 2001). ATP III identified 6 components of the metabolic syndrome that relate to CVD,  including abdominal obesity, atherogenic dyslipidaemia, raised blood pressure, insulin resistance ± glucose intolerance, pro‐inflammatory state, pro‐thrombotic state. Dyslipidemia (Appendix 1) represents a key pro‐atherogenic mechanism in HIV‐infected patients, and it  is typically attributed to the adverse effects of antiretroviral therapy (Calza 2008; Crook 2007; Morse 2006). The potential long‐term consequences of HAART‐associated hyperlipidaemia are not completely understood, but an increased risk of premature coronary artery disease has been reported in young HIV‐positive persons receiving HAART (Henry 1998; David 2002). Nonetheless,  the risk of cardiovascular events among HIV‐infected patients needs to be balanced against the remarkable benefits from HAART in terms of improvement in immune function and related morbidity and mortality (Bradbury 2008).

Dyslipidemia now is recognized as a significant potential adverse event in HIV‐infected patients who are receiving antiretroviral therapy (Anuurad 2009). The most compelling evidence that dyslipidaemia in HIV patients may increase the risk of myocardial infarction comes from the Data Collection of Adverse Events of Anti‐HIV Drugs study ( Friis‐Møller 2003). In this prospective, observational study, the relative risk of myocardial infarction attributed to antiretroviral therapy increased by 26% per year. Despite such limitations as short follow‐up, this finding suggests that dyslipidaemia in HIV patients should not be disregarded (Aberg 2009). Current guidelines support managing dyslipidaemia in HIV‐infected persons as in the general population (Dubé 2003). 

Description of the condition

The current standard of care for HIV treatment is a three‐drug regimen containing either a non‐nucleoside reverse transcriptase inhibitor ( NNRTI) or a protease inhibitor (PI) plus a ‘backbone’ consisting of two nucleoside/nucleotide‐reverse transcriptase inhibitors (NRTIs/NtRTIs) (DHHS 2009; Lundgren 2008; WHO 2010).

A clinical syndrome and  somatic changes (lypodistrophy), which are part of the metabolic syndrome associated to HAART in HIV patients have been extensively reported in subjects treated with PIs and NRTIs (Flexner 1998; Dubé 2003). Morphologic and metabolic alterations associated with HAART include isolated metabolic alteration, such as hyperlipidaemia, insulin resistance and hyperglycaemia,  and fat redistribution syndrome or lipodystrophy, which  may present as lipoatrophy, fat accumulation or both.

Though HIV infections has been associated with dyslipidaemia independent of antiretroviral therapy ( Riddler 2003), dyslipidaemia has been described as being more common and more severe in HIV patients receiving antiretroviral therapy than in patients not on therapy (Carr 1999). The severity of the dyslipidaemia and the typical pattern of the lipid profile differ among and within the classes of antiretroviral agents (Fontas 2004; Aberg 2009).

The dyslipidaemia can occur with or without central fat accumulation, peripheral fat loss, and inflammatory states. In patients with HIV, the metabolic syndrome has many of the same features as in HIV‐negative patients, being associated with an increased risk of cardiovascular and identified by the National Cholesterol Education Program (NCEP) as a secondary target for prevention of cardiovascular events (NCPE 2001). By contrast, reduction of serum low‐density cholesterol LDL cholesterol (LDL‐c) is the primary target of prevention of cardiovascular events.

Description of the intervention

Current recommendations suggest that HIV‐infected persons undergo evaluation and treatment on the basis of the Third National Cholesterol Education Program Expert Panel on Detection, Evaluation and Treatment of High Blood Cholesterol in Adults (NCEP ATP III) guidelines for dyslipidaemia, with particular attention to potential drug interactions with antiretroviral agents and maintenance of virologic control of HIV infection (Dubé 2003; NCPE 2001). While a hypolipidaemic diet and physical activity may certainly improve dyslipidaemia, pharmacological treatment becomes indispensable when serum lipid are excessively high for a long time or the patient has a high cardiovascular risk, since the suspension or change of an effective antiretroviral therapy is not recommended (da Silva 2009). Because there is a significant possibility for drug interaction of some lipid‐lowering agents with antiretroviral drugs (in particular statins are metabolized through the cytochrome (CYP) P450 system), care should be given to the choice of lipid‐lowering agents (LLAs).

The LLAs commonly used to treat lipid disorders in the general population are 3‐hydroxy‐3‐methyl‐glutaryl‐CoA reductase (HMG‐CoA reductase) inhibitors (statins), fibric acid derivatives (fibrates such as gemfibrozil, fenofibrate, clofibrate), bile acid binding resins (cholestyramine, colestipol, colesevelam), niacin, and inhibitors of intestinal absorption such as ezetimibe (Tungsiripat 2005). Of these, statins are the first‐line treatment if the primary problem is elevated LDL‐C or non‐HDL cholesterol levels and with triglyceride levels below 500 mg/dL, as they have been shown to reduce the progression of coronary artery stenosis and to reduce the risk of subsequent myocardial events (Gould 1998). Statins should also be the first‐line treatment in HIV patients with increased LDL‐C or non‐HDL‐C levels and with triglyceride levels below 500 mg/dL (Tungsiripat 2005). The HIV Medical Association of the Infectious Disease Society of America and the AIDS Clinical Trials Group advocate using statins for initial therapy if the primary problem is elevated LDL‐C and non‐HDL cholesterol (Dubé 2003). When the primary risk factor is elevated triglycerides, they recommend starting with a fibrates (Dubé 2003; NCPE 2001). Ezetimibe, a new cholesterol absorption inhibitor, lacks of side effects and P‐450 interactions and represents a potentially promising agent for use in patients with HIV infection (Dubé 2003; Tungsiripat 2005). Omega‐3 fatty acids (fish oils) alone or in combination reduce fasting serum triglyceride (Gerber 2008; Tungsiripat 2005; Wohl 2005). Bile acid binding resins are not recommended because they can interfere with the absorption of antiretroviral drugs (Tungsiripat 2005).

How the intervention might work

Dietary and lifestyle modifications are extremely important in the management of hyperlipidaemic patients. These non‐pharmacologic interventions include reduced intake of saturated fat and cholesterol, weight control, increased physical activity. Pharmacologic interventions become necessary when lifestyle changes alone prove inadequate. However, it is important that lifestyle changes be continued after the initiation of drug therapy to maximize the effectiveness of the pharmacologic intervention. The beneficial effects of statins are the result of their capacity to reduce cholesterol biosynthesis, as well as to the modulation of lipid metabolism, derived from their effect of inhibition upon HMG‐CoA reductase (Stancu 2001). Statins have anti‐atherosclerotic effects, that positively correlate with the percent decrease in LDL cholesterol. In addition, they can exert anti‐atherosclerotic effects independently of their hypolipidaemic action. In vitro and in vivo studies suggest that statins have direct anti‐inflammatory, anti‐thrombotic and plaque‐stabilizing effects via a number of mechanisms.  Because the mevalonate metabolism generates a series of isoprenoids vital for different cellular functions, from cholesterol synthesis to the control of cell growth and differentiation, HMG‐CoA reductase inhibition has beneficial pleiotropic effects. In endothelial cells, these metabolic changes contribute to favorable effects on nitric oxide (NO) bioavailability. Given the essential role of NO in preserving vascular structure and function, this effect of statins is of considerable therapeutic importance (Mason 2004). Consequently, statins reduce significantly the incidence of coronary events, both in primary and secondary prevention, being the most efficient hypolipidaemic compounds that have reduced the rate of mortality in coronary patients.  

Bile acid binders  (e.g., cholestyramine) bind bile acids in the intestine, causing the acids to be excreted rather than used to make bile and causing the liver to remove more LDL cholesterol from the bloodstream to make bile (Denke 2003). Cholesterol absorption inhibitor  (e.g., ezetimibe) decreases cholesterol absorption in the small intestine. Fibric acid derivatives  (e.g.,  gemfibrozil) increase the breakdown of lipids and speed the removal of VLDL from the bloodstream, thus reducing triglycerides blood levels; fibrates are less optimal alternative agents for hypercholesterolaemia (Dubé 2003). Lipoprotein synthesis inhibitor (e.g., niacin) slows removal of HDL and lowers triglyceride levels; at high doses, decreases production rate of VLDL, which is used to synthesize LDL.

Endothelial dysfunction is a critical early event in the development of atherosclerotic lesions and it is associated with major cardiovascular risk factors such as arterial hypertension, dyslipidaemia and diabetes mellitus (Celermajer 1992, Shankar 2004). Moreover, endothelial dysfunction may be caused by HIV infection itself as well as treatment‐related effects of the antiretroviral agents used to treat HIV (Friis‐Møller 2003). The available evidence suggests that PIs may induce endothelial dysfunction via their effects on both lipid and glucose metabolism (Dubé 2003, Shankar 2004).

Why it is important to do this review

This review should be conducted for the following reasons:

First, it has been demonstrated that patients on HAART are at increased risk for developing metabolic abnormalities that include elevated levels of serum triglycerides and low‐density lipoprotein cholesterol and reduced levels of high‐density lipoprotein cholesterol (Bradbury 2008; Manuthu 2008; Dubé 2003).
Second, this dyslipidaemia is similar to that seen in the metabolic syndrome, raising the concern that HAART therapy also potentially increases the risk for cardiovascular complications (Kotler 2008).
Third, recent guidelines and expert opinion based on extensive clinical experience suggest that HIV‐infected persons who have hyperlipidaemia should be managed similarly to those without HIV infection in accordance with the National Cholesterol Education Program (Aberg 2009; Dubé 2003; NCPE 2001). According to the US‐based Adult AIDS Clinical Trial Group (ACTG study A5047) Cardiovascular Disease Focus Group, treatment with pravastatin, fluvastatin or atorvastatin is recommended for antiretroviral‐linked hypercholesterolaemia, while lovastatin and simvastatin should be avoided due to interactions with PIs or non‐nucleoside reverse transcriptase inhibitors (NNRTI) and the risk of skeletal muscle toxicity (Fichtenbaum 2002). Many studies that evaluate the effect of statins for the treatment of antiretroviral‐associated dyslipidaemia have shown only partial responses to such therapy, with total and LDL‐C values being reduced by just 25%. (Calza 2003)

Fourth, the clinical effectiveness and safety of the drugs use for treating dyslipidaemia in patients with HIV infection should be assessed using clinical outcomes rather than surrogate markers.

Objectives

To assess the clinical effectiveness and safety of statins, ezetimibe, fibrates, or fish oil for treating dyslipidaemia in HIV‐infected patients receiving highly active antiretroviral therapy. Clinical effectiveness will be measured in terms of prevention (primary and secondary) of cardiovascular events (Fatal or non‐fatal myocardial infarction, stroke and angina).

Methods

Criteria for considering studies for this review

Types of studies

Randomised controlled trials (RCTs) irrespective its publication status (trials may be unpublished or published as an article, an abstract, or a letter). No country and sample size limitations will be applied. We will include RCTs conducted in either a hospital or community setting, or both. No HAART regimen limitations will be applied.

Types of participants

Adults (≥ 18 years) HIV‐infected patients with dyslipidaemia receiving HAART.

Types of interventions

Intervention:

  1. Statins (atorvastatin, fluvastatin, pravastatin, rosuvastatin, lovastatin and simvastatin). No restriction by dose will be applied.

  2. Ezetimibe, alone or in association with statin

  3. Fibrates (gemfibrozil, fenofibrate, benzafibrate)

  4. Omega‐3 polyunsaturated fatty acids/fish Oil

Comparison:
Statin (with or without ezetimibe), fibrates, or fish oil versus standard care (Diet and physical exercise), or placebo or no medication.
Statin versus statin.
Statin versus any other drug for treating dyslipidaemia.
Statin versus any combination of comparison mentioned above.

Any other head‐to‐head comparison.

Types of outcome measures

Primary outcomes

  • Coronary heart disease (CHD), including myocardial infarction (fatal and non‐fatal) and angina.

  • Cerebrovascular disease: transient Ischaemic attack and stroke (ischemic and haemorrhagic).

  • Overall mortality.

Secondary outcomes

Changes at the end trial of total cholesterol.
Changes at the end trial of low‐density lipoprotein cholesterol.
Changes at the end trial of triglycerides.
Number of withdrawals.
Reasons of withdrawals: i.e. lack of efficacy, by adverse events requiring discontinuation of intervention, and any other.

Safety.

  • Adverse event: "any untoward medical occurrence that may present during treatment with a pharmaceutical product but which does not necessarily have a causal relationship with this treatment" (Nebeker 2004).

  • Adverse drug reaction "a response to a drug which is noxious and uninitiated and which occurs at doses normally used in man for prophylaxis, diagnosis, or therapy of disease, or for the modification of physiologic functions" (Nebeker 2004). Such as:

  1. HMG CoA reductase inhibitors (statins): myopathy, increased liver enzymes.

  2. Nicotinic acid: flushing, hyperglycemia, hyperuricemia (or gout), upper gastrointestinal distress, hepatotoxicity.

  3. Fibric acids: dyspepsia, gallstones, myopathy.

  4. Ezetimibe: headache, back pain, musculoskeletal pain.

Search methods for identification of studies

Electronic searches

We will search the following electronic databases to find reports of relevant randomised clinical trials:

  • Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library latest issue);

  • MEDLINE (1966 to date);

  • EMBASE (1980 to date);

  • LILACS (1982 to date).

  • African Index Medicus (http://indexmedicus.afro.who.int/

  • ISI web of science.

Searching other resources

We will search the following trial database for ongoing and unpublished trials:

  • The Clinical Trials Search Portal of the World Health Organization (http://apps.who.int/trialsearch/)

The following web sites will also be searched:

  • Conference proceedings:

The American Conference for the Treatment of HIV (ACTHIV) (http://www.acthiv.org/).
Conference on Retroviruses and Opportunistic Infections (http://www.retroconference.org).
HIV/AIDS Conferences Worldwide (http://www.conferencealerts.com/aids.htm).

  • www.scirus.com

We will also check the reference lists of all the trials identified by the above methods.

Appendix 2 shows search strategy in Medline which will be adapted to the other electronic databases.
No language restrictions will be applied.

Data collection and analysis

We intend to summarise data by standard Cochrane Collaboration methodologies (Higgins 2009).

Selection of studies

Two review authors (Arturo Martí‐Carvajal ‐ AMC‐ and Mario Cruciani ‐MC‐) will independently assess each reference identified by the searches to see if they meet the inclusion criteria. We plan to resolve any disagreements that may arise through discussion.

Data extraction and management

Two review authors (AMC and MC) will independently extract data from the selected trials using standardised data extraction form.

Assessment of risk of bias in included studies

Each review author will independently assess the risk of bias of each trial using a simple form and will follow the domain‐based evaluation as described in the Cochrane Handbook for Systematic Reviews of Interventions 5.0.2 (Higgins 2009). We will compare the assessment results and discuss any discrepancies between ourselves. We aim to achieve agreement on the final assessment for each criteria by discussion.

We will assess the following domains as 'Yes' (i.e. low risk of bias), 'Unclear' (uncertain risk of bias) or 'No' (i.e. high risk of bias):

  1. Randomisation

  2. Concealment of allocation

  3. Blinding (of participants, personnel and outcome assessors)

  4. Incomplete outcome data

  5. Selective outcome reporting

  6. Free of other bias (Gurusamy 2009; Ioannidis 2008a; Ioannidis 2008b).

We will use the following definitions:

Generation of the allocation sequence

  • Yes (low risk of bias), 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 will be considered as adequate if a person who was not otherwise involved in the recruitment of participants performed the procedure.

  • Unclear (uncertain risk of bias), if the trial was described as randomised, but the method used for the allocation sequence generation was not described.

  • No (high risk of bias), if a system involving dates, names, or admittance numbers was used for the allocation of patients.

Allocation concealment

  • Yes (low risk of bias), 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 (uncertain risk of bias), if the trial was described as randomised, but the method used to conceal the allocation was not described.

  • No (high risk of bias), if the allocation sequence was known to the investigators who assigned participants or if the study was quasi‐randomised.

Blinding (or masking)

We will assess each trial (as 'Yes' 'Unclear', 'No') with regard to the following levels of blinding:

  • blinding of clinician (person delivering treatment) to treatment allocation;

  • blinding of participant to treatment allocation;

  • blinding of outcome assessor to treatment allocation.

Incomplete outcome data

  • Yes (low risk of bias), 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 (uncertain risk of bias), the report gave the impression that there had been no dropouts or withdrawals, but this was not specifically stated.

  • No (high risk of bias), the number or reasons for dropouts and withdrawals were not described.

We will further examine the percentages of dropouts overall in each trial and per randomisation arm and we will evaluate whether intention‐to‐treat analysis has been performed or could be performed from the published information.

Were all randomised participants analysed in the group to which they were allocated? (ITT analysis)

Yes (low risk of bias): Specifically reported by authors that ITT was undertaken and this was confirmed on study assessment, or not stated but evident from study assessment that all randomised participants are reported/analysed in the group they were allocated to for the most important time point of outcome measurement (minus missing values) irrespective of non‐compliance and co‐interventions.

No (high risk of bias): Lack of ITT confirmed on study assessment (patients who were randomised were not included in the analysis because they did not receive the study intervention, they withdrew from the study or were not included because of protocol violation) regardless of whether ITT reported or not

‘As‐treated’ analysis done with substantial departure of the intervention received from that assigned at randomisation; potentially inappropriate application of simple imputation.

Unclear (uncertain risk of bias): Described as ITT analysis, but unable to confirm on study assessment, or not reported and unable to confirm by study assessment.

Selective outcome reporting

  • Yes, adequate, pre‐defined, or clinically relevant and reasonably expected outcomes are reported on.

  • Unclear, not all pre‐defined, or clinically relevant and reasonably expected outcomes are reported on or are not reported fully, or it is unclear whether data on these outcomes were recorded or not.

  • No, inadequate, one or more clinically relevant and reasonably expected outcomes were not reported on; data on these outcomes were likely to have been recorded.

Free of other bias (Baseline imbalance):

  • Yes (low risk of bias), the trial appears to be free of other components that could put it at risk of bias.

  • Unclear (uncertain risk of bias), the trial may or may not be free of other components that could put it at risk of bias.

  • No (high risk of bias), there are other factors in the trial that could put it at risk of bias, e.g., no sample size calculation made, early stopping, industry involvement, or an extreme baseline imbalance.

1.‐ Trials that achieve a 'yes' for generation of allocation sequence, adequate allocation concealment, adequate blinding, adequate handling of incomplete outcome data, no selective outcome reporting, and without other bias risks will be considered as low‐bias risk trials.

2.‐ Trials at high risk of bias are either 'No' or 'Unclear' on the majority of domains.

Measures of treatment effect

  • For binary outcomes, we will compare using the risk ratio (RR), and we will present each result with a 95% confidence intervals (CI): coronary heart disease, cerebrovascular disease, cardiac death, reasons for withdrawals and safety.

  • For continuous outcomes, we will compare using the difference in means, and we will present each result with a 95% CI for: number of withdrawals, total cholesterol, low‐density lipoprotein cholesterol, and triglycerides.

Unit of analysis issues

The unit of analysis will be the included randomised clinical trials considering the studied patients. A single measurement for each outcome from each participant will be collected and analysed.

Dealing with missing data

We will assess the percentages of dropouts overall for each included trial and per each randomisation arm and we will evaluate whether an intention to treat analysis has been performed or could be performed with the available published information.

In order to allow us to undertake an intention‐to‐treat analysis, we will seek data from the trial authors on the number of participants by treatment group, irrespective of compliance and whether or not the participant was later thought to be ineligible or otherwise excluded from treatment or follow up.

Assessment of heterogeneity

We will quantify statistical heterogeneity using the I2 statistic, which describes the percentage of total variation across trials that is due to heterogeneity rather than sampling error (Higgins 2003). We will consider there to be significant statistical heterogeneity if I2 >50% (Higgins 2009).

Assessment of reporting biases

Where we suspect reporting bias (see ‘Selective reporting bias’ above), we will attempt to contact study authors asking them to provide missing outcome data. Where this is not possible, and the missing data are thought to introduce serious bias, the impact of including such studies in the overall assessment of results will be explored by a sensitivity analysis. 

We will also attempt to assess whether the review is subject to publication bias by using a funnel plot to graphically illustrate variability between trials. If asymmetry is detected, causes other than publication bias will be explored (Selective outcome reporting, poor methodological quality in smaller studies, true heterogeneity) (Higgins 2009). Funnel plot will be conducted if ten or more randomised controlled trials (RCTs) are included.

Data synthesis

If the eligible trials are sufficiently comparable (I2 <50%), we will summarize their findings using a random‐effect model. We will attempt to explain the cause of the heterogeneity according the Cochrane Handbook of Systematic Reviews for Interventions (Higgins 2009).

Subgroup analysis and investigation of heterogeneity

We anticipate clinical heterogeneity in the effect of the intervention and we propose to conduct the following sub‐group analyses:
1.‐ By HAART regimen.
2.‐ By intervention.
3.‐ By country.
4.‐ By gender.

Subgroup analysis will be only performed for primary outcomes.

Sources of heterogeneity in the assessment of the primary outcome measure will be explored by subgroup analyses and meta‐regression analyses. The meta‐regression analyses will assess the effect of methodological quality (High versus low), statin type, HAART regimen, and country. Meta‐regression will be only conducted if >10 RCTs are included.

Sensitivity analysis

If sufficient trials are identified, we plan to conduct a sensitivity analysis comparing the results using all trials as follow:

1.‐ Those RCTs with high methodological quality (studies classified as having a 'low risk of bias' versus those identified as having a 'high risk of bias') (Higgins 2009).

2.‐ The effect of placebo control.

We will also evaluate the risk of attrition bias, as estimated by the percentage of participants lost. Trials with a total attrition of more than 30% or where differences between the groups exceed 10%, or both, will be excluded from meta‐analysis but will be included in the review.

Summary of findings tables

We will use the principles of the GRADE system (Guyatt 2008) to assess the quality of the body of evidence associated with specific outcomes (coronary heart disease, cerebrovascular disease, cardiac death, safety, number and reasons of withdrawals) in our review and construct a Summary of Findings (SoF) table using the GRADE software. The GRADE approach appraises the quality of a body of evidence based on the extent to which one can be confident that an estimate of effect or association reflects the item being assessed. The quality of a body of evidence for considers within study risk of bias (methodologic quality), the directness of the evidence, heterogeneity of the data, precision of effect estimates and risk of publication bias.