Full Title: Safety of Probiotics Used to Reduce Risk and Prevent or Treat Disease
Expected Release Date: late 2009
Contents
Background
Project Purpose
Key Questions
Methods
Analysis
Rating the Strength of the Evidence
Report
References
Background
Probiotics are defined as dietary
supplements of live bacteria or yeasts thought to be healthy for the host
organism. According to the definition currently adopted by the Food and
Agriculture Organization of the United Nations & World Health Organization (FAO/WHO),
probiotics are live microorganisms that when administered in adequate amounts
confer a health benefit on the host (FAO/WHO Expert Consultation, 2001). Prebiotics are a category of functional foods, defined as non-digestible food ingredients
that beneficially affect the host by selectively stimulating the growth and/or
activity of one or a limited number of bacteria in the colon and thus improve
host health. Synbiotics refer to nutritional supplements combining
probiotics and prebiotics to form a synbiotic relationship.
Probiotics can be formulated into
many different types of products, including foods, drugs, and dietary
supplements. Most commonly used as probiotics are species of Lactobacillus and
Bifidobacterium but there are a number of other probiotic genera such as Saccharomyces,
Streptococcus, Enterococcus, and Bacillus. Lactic acid bacteria have been used
for preservation of food by fermentation for thousands of years (World
Gastroenterology Organisation, 2008).
There is a growing awareness of potential
safety concerns such as the possibility of gene transfer and isolation of
probiotic bacteria from infection sites; and the pathogenicity, infectivity,
toxicity, and intrinsic properties of the bacteria may have to be studied more
closely (Ishibashi & Yamazaki, 2007). However, to date there is no
systematic review synthesizing the available evidence of symptomatic health
outcomes in patients.
Return to Contents
Project Purpose
The purpose of the project is to
catalogue what is known about the safety of probiotics (Lactobacillus, Bifidobacterium, Saccharomyces, Streptococcus, Enterococcus, and Bacillus)
used in research to reduce the risk of, prevent, or treat disease. The literature
review will assess the quality and completeness of the available information,
and our confidence to interpret this information. The project will provide information
relevant to practitioners, researchers, and regulators for assessing the value
and safety of probiotic administration to reduce risk, or prevent or treat
disease, as well as to identify priorities or needs for future research.
Return to Contents
Key Questions
Note: Questions #1 and #2 relate
to six taxonomic groups (Lactobacillus, Bifidobacterium, Saccharomyces, Streptococcus, Enterococcus, and Bacillus).
- What is the evidence that the
active (e.g., live or viable) and lyophilized forms of probiotics (Lactobacillus, Bifidobacterium, Saccharomyces, Streptococcus,
Enterococcus, and Bacillus) as single ingredients or in
combination with other probiotics or prebiotics in all delivery vehicles
(and formulations) when used to cure, treat, mitigate or prevent a disease
or reduce disease risk are safe in the short-term? Long-term?
- What safety parameters are
collected in clinical studies (Phases I-IV)?
- What harms are reported in
clinical studies (Phases I-IV)?
- What harms are reported in
case reports?
- What safety parameters are
collected in population surveillance studies and other observational
studies, and do these include only standard clinical safety parameters
(e.g., standard blood chemistry profiles) or also expanded laboratory or
clinical testing unique to the use of probiotics?
- What harms are reported in
population surveillance studies and other observational studies?
- What harms are reported in
human mechanistic studies?
- Do the studies describe an
antibiotic therapy designed to treat unintended pathology caused by the
administered organism?
- Do the studies describe
methods for recovery of the administered organism from either the
gastrointestinal tract or serum?
Note: All subsequent
questions refer only to those taxonomic groups for which harms are reported in
Question #1.
- What are characteristics and
associations of the reported harms in Question 1?
- What interactions between
probiotics and medications are reported?
- What harms related to acquired
antibiotic resistance and/or transferability are reported?
- What is the nature of harms
(e.g., toxicogenic,
immunologic, hematologic, deleterious physiologic or metabolic activity,
allergic, blood infections, hematocytometric values, liver and renal
function enterotoxin, production, proteases, or opportunistic infection, etc.),
and do these include only standard harms or also harms that might be
uniquely applicable to the use of a probiotic?
Note: All subsequent
questions refer to the literature reporting harms described in Questions
#1 and #2, but are primarily limited to the literature examining the
following four taxonomic groups (Lactobacillus, Bifidobacterium, Saccharomyces, and Streptococcus). Depending on the search volume
and available resources Enterococcus and Bacillus studies will also be
considered.
- What is the evidence that harms
of Lactobacillus, Bifidobacterium, Saccharomyces, and Streptococcus identified in Question #1 differ by product and
delivery characteristics?
- What is the scientific
evidence that harms differ by delivery vehicle including excipients or
novel delivery vehicles?
- What is the scientific
evidence that harms differ by genus, species, and strain (including
intraspecies strain variations)?
- What is the scientific
evidence that harms differ between active and lyophilized forms of
probiotics?
- Does harm differ by products
containing a single probiotic vs. a mixture of probiotics?
- Does harm differ by products
containing only probiotics and those containing a mixture of probiotics
and prebiotics?
- How do the harms of Lactobacillus, Bifidobacterium, Saccharomyces, and Streptococcus vary based on (a) dose (cfu); (b) timing; (c) mode of administration
(e.g., catheter); (d) age (all ages, including infants), gender,
ethnicity, disease or immunologic status of the patient; (e) relationship
to efficacy?
- Is there a threshold or
dose-response relationship between probiotics and harm? Does the
duration of intervention relate to harm?
Note: Dose (cfu) of active microorganisms needs to be
verified. When viable (active) probiotics are added to a vehicle,
their viability and/or dose may be compromised. Literature reports should
include verification of dose if available. If not, the Evidence-Based
Practice Center may try to verify the dose with investigators.
- Is there a relationship between
time of onset of harm and time of probiotic administration (e.g., prior
to onset of disease under study, after disease onset)? How does
time of exposure affect harm? Is harm sustained after the
intervention or exposure stops?
- Does the route of
administration (e.g., orally, jejunostomy tube, central venous catheter)
relate to harm?
- How does harm relate to
subpopulations, including different age groups (specifically including
neonates and infants under age 24 months), men and women, ethnic/race
subgroups, or health status (healthy to high risk) individuals?
- Do randomized controlled studies
that report harm show efficacy or no efficacy?
- How often does harm associated
with Lactobacillus, Bifidobacterium, Saccharomyces,
and Streptococcus lead to hospital admission or lengthened
hospitalization?
- How does harm associated with Lactobacillus, Bifidobacterium, Saccharomyces, and Streptococcus relate to use of concomitant antibiotics, confounding diet therapies,
corticosteroid use, immune suppressants, or other potential confounders?
Note: Probiotics may be
unreliable in content. For example, ConsumerLab.com analyzed 25 probiotic
products only to find that the number of live organisms identified did not come
close to the manufacturers' claims. United States and foreign studies that
independently analyze commercial probiotic products demonstrate that lower microorganism
counts exist than the manufacturers claim and that products may not even
contain the organisms that the manufacturers claim.
In addition, taxonomy and
nomenclature may not be standardized. There may be a difference in
taxonomy and nomenclature used in research vs. commercial use, as well as
between countries. National and international organizations recommend
that probiotics be named according to rules described in the International Code
of Nomenclature of Bacteria to ensure understanding on an international basis
and urge that for the sake of full disclosure, probiotic strains be deposited
in an internationally recognized culture collection.
Therefore, the evidence review will
take into consideration the quality of the reporting of product
characteristics. The review aims to document the type of microorganism
(genus, species and strain) using a standardized taxonomy and nomenclature,
potency (number of viable bacteria per dose), purity (presence of contaminating
or ineffective bacteria), and in which culture collection the test
microorganism is deposited.
Return to Contents
Methods
Electronic search for literature review
Preliminary searches have shown
that there is a high volume of literature that will need to be screened
carefully for potentially relevant articles on the safety of probiotics. These
prior searches also suggest adding the term 'prebiotics' to the search strategy
as they seem to identify additional relevant studies that should undergo
further scrutiny. In addition, the term 'synbiotic' will be used in the search
strategy to adequately reflect the scope of this project.
There are currently about 5000
publications on probiotics, prebiotics and synbiotics indexed in PubMed®. This
suggests that the final search outcome will be high, presumably identifying 10,000
records. This high volume is mainly a result of the difficulty of reliably
identifying adverse events. While search filters exist for effectiveness
studies, filters to address harms are not successful in identifying relevant
studies.
The searches will be performed without restriction to
publication year. The search will not be restricted to the investigated probiotic
genera in order as not to miss studies that do not mention the genus in the
title, abstract or keywords of the publication. The genera terms alone (without
reference to the use as probiotics) are not useful search terms as they add a
very high number of irrelevant publications (e.g., all studies on Bacillus
infections). Given that there is a large number of prebiotic, probiotic, and synbiotic products marketed as dietary supplements, food or drugs the search will not use
product names; and an incomplete list may introduce bias in the selection of
studies under review. The electronic search will screen out studies that are
indexed by the individual databases as studies in animals where possible.
The literature will be searched
without language restrictions, taking also into consideration that a
substantial proportion of research is published in Asian language publications.
While uncertainty exists regarding whether the strains investigated in these
studies are similar to those common in the US market, these studies need to be
assessed. Where necessary, we will attempt to contact authors to identify
strains tested as well as any other information needed to answer the questions.
The following databases will be
searched as sources for safety data on probiotics:
Databases
- DARE (Database of Abstracts of Reviews of Effects).
- Cochrane library of systematic reviews.
- CENTRAL (Cochrane Central Register of Controlled Trials).
- PubMed (National Library of Medicine, includes MEDLINE®).
- EMBASE (Biomedical and pharmacological bibliographic
database).
- CINAHL® (Cumulative Index to Nursing and Allied Health
Literature).
- AMED (Allied and Complementary
Medicine).
- MANTIS (Manual, Alternative and Natural Therapy Index
System).
- TOXLINE (biochemical, pharmacological, physiological, and
toxicological effects of drugs and other chemicals).
- TOXFILE (biochemical, pharmacological, physiological, and
toxicological effects of drugs and other chemicals).
- NTIS (National Technical Information Service).
- AGRICOLA (agricultural journals).
|
Box 1: PubMed® search strategy
PubMed – 1966-2009
probiotic* OR prebiotic* OR
pre-biotic* OR synbiotic*
NOT
animals NOT humans
|
Other sources
- Hand search International Journal of Probiotics and
Prebiotics.
- Clinicaltrials.gov (about 80 completed probiotic trial
registered; will be searched during planned update searches).
- References of included studies.
- References of relevant reviews.
- Personal files from related topic projects.
- The usefulness for the identification of safety data from
searching MedWatch, the Web pages of Center for Food Safety and Applied Nutrition (CFSAN), Food and Drug Administration (FDA) and Center for Drug Evaluation and Research (CDER), and searching CAERS (CFSAN Adverse Event
Reporting System) will be explored.
In addition, manufacturers
producing probiotic, probiotic, and synbiotic products will be identified.
Inclusion
screening
Inclusion criteria
- Participants:
- Studies in human participants are eligible for inclusion in the review, animal and in vitro studies are excluded.
- Intervention:
- Studies using probiotics or synbiotics to cure, treat, mitigate or prevent a disease or reduce disease risk (including probiotic drinks or supplements 'to boost immunity' or similar) are eligible for inclusion in the review.
- Comparator and Study design:
- Primary data studies will be considered without study design restriction but case studies and case series will only be considered if they explicitly address the effect of probiotics or synbiotics intake:
- Randomized controlled trials (RCTs), controlled clinical trials (CCTs), and Cohort studies with 2 arms comparing the use of probiotics or synbiotics to placebo, other treatment or other types of probiotics/prebiotics/synbiotics.
- Before-after studies and time series with measurements before and after introducing probiotics or synbiotics.
- Case series (no comparator) that address the effects of probiotics or synbiotics.
- Case reports that explicitly address the effects of probiotics or synbiotics.
- Mechanistic probiotics or synbiotics studies addressing patient health outcomes.
- Case-control studies that focus on probiotics or synbiotics as predictors of an adverse event in patients.
- Outcomes:
- Studies that address harms such as adverse patient health outcomes, in particular symptomatic outcomes, are included in the review; studies that only report intermediate outcomes such as gene transfer or gastric colonization with unclear relevance to patients' health are not eligible for inclusion in the review.
- Genus:
- Studies investigating Lactobacillus, Bifidobacterium, Saccharomyces, Streptococcus, Enterococcus, and/or Bacillus alone or in combination with other ingredients are eligible for inclusion in the review. Studies where it cannot be verified what group of probiotics was consumed are excluded.
Title and Abstract Inclusion screening
The initial relevance screening
will be performed using the reference manager software Endnote. Endnote allows
the importing of titles, abstracts, and keywords for each reference identified
through electronic searches. All identified records will be screened
independently by two reviewers in order as not to miss potentially relevant
studies. Records deemed potentially reporting safety information by at least
one reviewer will be ordered as full text copies for further scrutiny.
Identifying safety data is
challenging since most publications focus on the clinical effectiveness of the
intervention in question with either no, sparse or incomplete and
non-systematic reporting of safety aspects. The review team will follow
inclusive decision rules for ordering full paper copies of publications in
order as not to miss studies that report on harms in the full publication but
do not indicate so in the title, abstract or keywords of the publication.
Full Text Inclusion Screening
Two reviewers will independently
screen the selected full text publication using a standardized form outlining
the inclusion criteria. Any disagreement will be resolved through discussion,
through consultation with the review team, or with other input such as the
local content expert or the technical expert panel (TEP).
The inclusion screening process
will also identify all randomized trials (RCTs) reporting patient health
outcomes in human participants using probiotics or synbiotics of the genus Lactobacillus,
Bifidobacterium, Saccharomyces, Streptococcus, Enterococcus, or Bacillus to
cure, treat, mitigate or prevent a disease or reduce disease risk compared to
placebo, another probiotic, prebiotic or synbiotic, other, or no intervention
in order to have a denominator for safety-addressing RCTs.
The report will document the literature flow in a standard
flow diagram to ensure transparency of the review conduct. The report will
include a table with studies that underwent full text inclusion screening but
which did not meet inclusion criteria and will list the reason for exclusion.
Data Abstraction and Quality Assessment
The following outlines the type
of information that will be abstracted for the included studies. It will
include information about the type of study; the participants; the product
containing probiotics or synbiotics; the assessed harms and adverse events; and
the results of the study regarding the safety of the intervention. The data
abstraction will consist of free text abstraction and categorization of
information using defined categories where possible and appropriate.
The data abstraction form will be
further developed using a sample of included studies with input from the research
team and TEP.
|
Box 2: Draft Data Abstraction Outline [not
final form]
Study details:
ID, Author, Year, Number of
publications, IDs of multiple publications, Country, Study Design, Main aim of
study assessment of adverse events?
Participant information:
Age category, Gender
distribution, Ethnicity, Disease or immunologic status, Distinct patient
subgroup; Concomitant antibiotic, Diet therapy, Corticoid steroid use, Immune
suppressant use, Other potential confounders reported.
Product details:
Product (free text), Product
type (single probiotic or mixed; probiotics plus prebiotics, synbiotics), Genus
(Lactobacillus, Bifidobacterium, Saccharomyces, Streptococcus, Enterococcus, or
Bacillus or mixture), Species according to International Code of Nomenclature
of Bacteria taxonomy, Strain including intra-species strain variation, Form
(e.g., active, lyophilized), Means of delivery (e.g., yoghurt, milk drink, pill,
other), Route of administration and mode of delivery (e.g., catheter), Dose of
product (exact dose, frequency of intake), Potency (dose of active
microorganisms, number of viable bacteria per dose), Purity (presence of
contaminating or ineffective bacteria), Test culture (in which culture
collection the test microorganism is deposited), Timing, Duration of exposure.
Safety parameters:
Assessed safety parameters (free
text), Assessment categorization—standard clinical safety parameters (e.g.,
standard blood chemistry profiles); Presence of expanded laboratory or clinical
testing unique to the use of probiotics extract, All reported harms—free text,
Categorization of harms (CTC classification system adaptation), Hospital admissions,
Treatment for administered organism—free text, Duration of symptoms after
probiotics exposure stopped.
Analysis details:
Differentiation of
probiotics and medication effects (e.g., multi-variate analysis),
Differentiation of probiotics and confounders, Probiotic-medication interaction
addressed?, Does the intervention appear to be effective? [RCTs only]
|
We will attempt to contact principal
investigators of primary studies to verify
the dose of viable probiotics where not stated in the original publication and
where applicable.
Each study will also be assessed
regarding its quality. We are considering a wide range of study designs in this
review and some quality dimensions will be specific to the individual study
design (e.g., concealment of treatment allocation (RCTs)), while others are
sources of bias that can apply to many study designs (e.g., blinding of outcome
assessors). The quality assessment will also incorporate the quality of the
reporting, of both the product and probiotic genus, species and strain; as well
as the methods and the reporting of the assessment and the documentation of
observed harms.
The data abstraction and quality
assessment will be performed in duplicate with two reviewers independently reviewing
the publications using a standardized form. The numerical results for the
eligible outcomes will be extracted by the EPC statistician. Any disagreement
will be resolved through discussion, through consultation with the review team,
or with other input such as from the local content expert or the TEP.
The study characteristics,
results, and the quality assessment will be documented in concise evidence
tables.
Return to Contents
Analysis
Most of the outlined review questions
will be answered by providing descriptive data (e.g., number of studies
reporting adverse events, type of harms, etc.).
Where appropriate, odds ratios of
harms for treatment and control arms will be computed for each study and pooled
across studies in a meta-analysis for a summary estimate. Studies will be
included for analysis if they include information about the total number
of people in each group as well as the number of people with events in each
group. In cases where the number of events are reported for one group
within a study but not for the other group, we assume that zero events occurred for
this second group.
For groups of events that appear
in at least three trials, a meta-analysis can estimate the odds
ratio and its 95% confidence interval. Since adverse events are
generally rare, conditional pooling using exact methods will provide a fixed
effects estimate of the odds-ratio. Analyses will be conducted with Stat Xact
Procs for SAS.
Subgroup analyses will likely be narrative
in order to be able to compare between study designs and other variables in the
heterogeneous dataset. Further input, e.g., about effect modifiers or pertinent
subgroups, will come from the local content expert and the TEP.
Multiple publications of the same
study will be noted but counted (and extracted, quality assessed and analyzed)
as one study to ensure that the same participants do not enter the analyses
multiple times. Multiple publications are defined by the investigated patients.
The proportion of RCTs that address
adverse events will also be determined relative to the total number of
identified RCTs reporting patient health outcomes in human participants using
probiotics or synbiotics of the genus Lactobacillus, Bifidobacterium, Saccharomyces,
Streptococcus, Enterococcus, or Bacillus to cure, treat, mitigate or prevent a
disease or reduce disease risk compared to placebo, another probiotic,
prebiotic or synbiotic, other or no intervention. This assessment will answer
the question how many high-evidence-level studies do and do not address the
safety of using probiotics or synbiotics.
Return to Contents
Rating the Strength of the Evidence
A synopsis of the evidence will be provided for each of the
key research questions. The body of evidence will be evaluated by taking the risk
of bias of individual studies, the consistency across studies, and where available
and appropriate, the directness and the precision of results into account as
outlined in the Methods Guide for Effectiveness and Comparative Effectiveness
Reviews (http://effectivehealthcare.ahrq.gov/repFiles/2007_10DraftMethodsGuide.pdf [PDF File, 1.1 MB; PDF Help]).
Return to Contents
Report
The review will be documented in
an evidence report. The report will document the key questions, the inclusion
criteria, the literature search strategy, the literature flow and a list of
company names and/or contacts for all manufacturers or corporate entities that
produce probiotics products. The report will document eligible studies in
evidence tables showing individual study characteristics, results, and the
critical appraisal of each study. The analyses will be presented in tables and
figures. The report will include an evaluation of the overall body of evidence
available for the key research questions. Apparent gaps in the evidence base
will be outlined to facilitate the identification of future research
priorities.
The review will be conducted with
usual oversight under AHRQ EPC processes, and the draft report will be reviewed
by all project members, the local content expert, the members of the technical
expert panel, and additional peer reviewers to ensure a transparent, reliable,
and valid evaluation of the existing research on the safety of probiotics.
Return to Contents
References
Agency
for Healthcare Research and Quality (2007). Methods Reference Guide for
Effectiveness and Comparative Effectiveness Reviews. Version 1.0. [Draft posted
Oct. 2007] Rockville, MD. Available at: http://effectivehealthcare.ahrq.gov/repFiles/2007_10DraftMethodsGuide.pdf (PDF File, 1.1 MB; PDF Help).
FOA/WHO Expert Consultation
(2001). Health and nutritional properties of probiotics in food including
powder milk with live lactic acid bacteria. Report of a Joint Food and
Agriculture Organization of the United Nations & World Health Organization
Expert Consultation on Evaluation of Health and Nutritional Properties of
Probiotics in Food Including Powder Milk with Live Lactic Acid Bacteria. Cordoba, Argentina, 1-4 October 2001.
Ishibashi N, Yamazaki S.
(2001). Probiotics and safety. American Journal of Clinical Nutrition 72(2
Suppl.), 465S-470S.
World Gastroenterology Organisation (2008). World
Gastroenterology Organisation Practice Guideline: Probiotics and prebiotics.
WGO May 2008.
Return to Contents
Current as of September 2009
Internet Citation:
Safety of Probiotics Used to Reduce Risk and Prevent or Treat Disease, Review Protocol. Agency for Healthcare Research and Quality, Rockville, MD. http://www.ahrq.gov/clinic/tp/probiotictp.htm
540 Gaither Road Rockville, MD 20850