CommunityBased Studies of Consumer Antiseptics FDA Nonprescription Drug Advisory Committee Meeting

1
Community-Based Studies of Consumer
AntisepticsFDA Nonprescription Drug Advisory
Committee Meeting

• Allison E. Aiello, PhD, MS
• Assistant Professor of Epidemiology
• University of Michigan-School of Public Health
• Center for Social Epidemiology Population Health

2
Outline

• Literature Review
• Methodological Issues
• Summary and Future Research Needs

3
Overall Goal

• Estimate the reduction in risk attributed to
  specific hand hygiene products
  
• Plain soap handwash
• Antiseptic soap handwash
• Triclosan,Triclocarban
• Hand sanitizers
• Alcohol-based
• Non-alcohol based hand sanitizers
• Benzalkonium chloride

4
Methods

• Use of earlier systematic review articles
• Aiello, AE Larson, EL. (2002) What is the
  evidence for a causal link between hygiene and
  infections? Lancet Infect Dis. Vol.2
  
• Curtis, V Cairncross, S. (2003) Effect of
  washing hands with soap on diarrhoea risk in the
  community a systematic review. Lancet Infect
  Dis. Vol.3
• Meadows, E Le Saux, N. (2004) A systematic
  review of the effectiveness of antimicrobial
  rinse-tree hand sanitizers for prevention of
  illness-related absenteeism in elementary school
  children. BMC Pub Health.(4)50
• PubMed database search for other articles from
  1980 to 2005
  
• Key word combinations hygiene, infection, soap,
  washing, handwashing, community, alcohol,
  antibacterial, triclosan, day care, school, hand
  sanitizer

5
Methods Continued

• Inclusion Criteria
• Outcome
• Culture confirmed infection, symptoms of
  infection, or absenteeism associated with
  infectious illness
  
• Study Design
• Community based
• Intervention or cross-over study with or without
  formal randomization
  
• Intervention arm provided
• Plain soap (not identified)
• Antiseptic soap
• Alcohol-based hand sanitizer
• Non-alcohol hand sanitizer

6
Number of Studies Meeting Criteria

• Soap (plain/unidentified) N8
• Antiseptic Soap N5
• Alcohol-Based Hand Sanitizers N9
• Non-Alcohol-Based Hand Sanitizers N2

7
Plain or Unidentified Soap Versus Control (N8)
8
Plain or Unidentified Soap (N8) Study Methods
Overview

• Formal randomization (n3)
• Soap form
• Bars (n8)
• Reported plain or non-medicated soap (n4)
• Unknown soap type (n4)
• Educational component
• Wash hands at critical points (n7)
• Hygiene promotion seminar (n3)
• Washing of dishes (n1)
• Follow regular routine (n1)

9
Plain or Unidentified Soap (N8) Study Methods
Overview

• Outcomes
• Diarrhea Incidence/Risk (n6)
• Prevalence, Healthcare Visits
• Culture confirmed Shigella sp.
• Impetigo (n2)
• Skin and eye disease (n1)
• Runny nose, cough, and pneumonia (n1)

10
Plain or Unidentified Soap (N8) Study Methods
Overview

• Exposure measurements
• Measured Soap use or Hand Washing (n6)
• Controlled for confounding or reported balance on
  covariates (n7)
  

11
Plain or Unidentified Soap (N8)ResultsReductio
n in Diarrhea Incidence (n6)
All Results Statistically Significant (P 12
Plain or Unidentified Soap (N8) Results
Continued

• Other significant findings
• Luby, S. et al. (2005) Lancet
• 51 reduction in cough or difficulty breathing
• 54 reduction in coryza (runny nose)
• 50 reduction in pneumonia
• 34 reduction in impetigo
• Null findings
• Luby, S. et al. (2002) Am J Trop Med Hyg
• 25 reduction in total impetigo (p0.21)
• No prompting of change in hygiene

13
Plain or Unidentified Soap Type (N8) Summary
of Results

• Reduction in incidence of diarrhea ranged from
  30 to 89
  
• Median reduction 53
• Similar reduction range for other outcomes

14
Plain or Unidentified Soap Type Study Design
Issues

• Most studies prior to 2004 lack formal
  randomization
  
• Not possible to mask participants or
  interviewers
  
• Only two studies used techniques to control for
  clustering
  
• Limited measurement on hygiene / soap use

• Varying definitions of symptoms and reporting
  methods across studies
  
• All studies were conducted outside of USA
• High risk populations
• Difficult to tease apart the effect of soap use
  from that of hygiene education
  
• Only one study asked participants to follow
  regular routine
  

15
Plain or Unidentified Soap Type (N8)
Conclusions

• Consistent reductions observed
• regardless of varying methods and study
  populations
  
• Plain soap in conjunction with proper hygiene is
  effective in reducing diarrheal illness in high
  risk populations

16
Antiseptic Soap Versus Plain Soap or Control (N5)
17
Antiseptic Soap (N5) Study Methods Overview

• Comparison Groups
• Antiseptic Soap versus Plain Soap (placebo)
  (n2)
  
• Antiseptic Soap versus Plain Soap (placebo) and
  versus Control Group (nothing) (n3)
  
• Formal randomization (n3)
• Soap form
• Bars (n4) (1.2 triclocarban)
• Liquid (n1) (0.2 triclosan)
• Educational component
• Washing hands at critical points (n3)
• Follow regular routine (n2)

18
Antiseptic Soap (N5) Study Methods Overview

• Outcomes
• Diarrhea Incidence/Risk (n3)
• Impetigo (n2)
• Pneumonia (n1)
• Symptoms of infection (n2)

19
Antiseptic Soap (N5) Study Methods Overview

• Measured Soap use and/or Hand Washing (n4)
• Masked participants and interviewers (n4)
• Among placebo groups only
• Controlled for confounding or reported balance on
  covariates (n5)
  

20
Antiseptic Soap vs. Plain Soap Reduction in
Diarrhea Incidence (n2)
10 Reduction
Luby (2004) Pakistan, Children
Larson (2004) United States, Families
6 Higher
Not statistically significant
Not statistically significant and not designed
or powered to make this comparison
21
Antiseptic Soap vs. Plain SoapReduction in
Incidence of Other Symptoms / Infections (n3)
No results were statistically significant
22
Antiseptic Soap versus Control Group (No Soap
Provided in Control Group) Reduction in
Incidence of Impetigo and Diarrhea (n3)
All results statistically significant (P 23
Antiseptic Soap Summary of Results

• Antiseptic Soap vs. Plain Soap (n3)
• No statistically significant differences for all
  infectious symptoms
  
• Antiseptic Soap vs. Control Group (no soap)
  (n2)
  
• Reduction in incidence of diarrhea ranged from
  29 to 50
  
• Similar to reductions associated with use of
  plain soap
  

24
Antiseptic Soap Study Design Issues

• Study Design Limitations
• Possible viral or parasitic etiology for symptom
  definitions
  
• High risk groups
• Study Design Strengths
• All studies used techniques to control for
  clustering
  
• All studies measured baseline hygiene
  information
  
• Hygiene practices over duration of the study
  (n1)
  
• Product monitoring (n2)
• Three studies masked participants and/or
  interviewers
  
• Use of a placebo plain soap (n3)
• Extensive follow-up for symptoms

25
Antiseptic Soap Conclusions

• Lack of evidence that antiseptic soaps provide a
  benefit beyond plain soap in community setting in
  US and Pakistan
  
• Diarrhea
• Impetigo
• Other infectious symptoms
• When compared to a control group (no provision of
  plain soap or hygiene education)
  
• Antiseptic soap with hygiene education is an
  effective intervention for reducing impetigo and
  diarrheal illness in high risk groups

26
Alcohol-Based Hand Sanitizers (N9)
27
Alcohol-Based Hand Sanitizer (N9) Study
Methods Overview

• Comparison Groups
• Alcohol Education versus Control Group
  (nothing) (n6)
  
• Alcohol Education versus Control Group
  (education) (n2)
  
• Alcohol versus Control Group (nothing) (n1)
• Formal randomization (n4)
• Cross-over (n2)
• Alcohol form
• Hand rinse (n1)
• 60 isopropyl alcohol
• Disinfectant (n2)
• 63 isopropyl alcohol, 70-90 ethanol and
  isopropanol
  
• Foam
• Instant hand sanitizer (n5)
• 60-62 or unknown

28
Alcohol-Based Hand Sanitizer (N9) Study Methods
Overview

• Outcomes
• Gastrointestinal illness (n3)
• Upper respiratory illness (n3)
• Viral Respiratory Illnesses (n1)
• Symptoms of infection (n3)
• Absence related

29
Alcohol-Based Hand Sanitizer (N9) Study
Methods Overview

• Measured Alcohol use (n3)
• Supplies (n3)
• Frequency of use (n1)
• Total hand hygiene practices (n1)
• None masked participants or interviewers
• One study collected illness data from masked
  parents of children in the study
  
• Controlled for confounding or reported balance on
  covariates (n4)
  

30
Alcohol-based Hand Sanitizer Reduction in
Diarrheal Illness (n3)
All P Reduction
31
Alcohol-based Hand Sanitizer Reduction in
Respiratory Illness (n4)
P Not statistically significant
32
Alcohol-based Hand Sanitizer Symptoms of
Infection (n4)




All P reduction
33
Alcohol-based Hand Sanitizer Summary of Results

• Reductions in diarrheal illness ranged from 48
  to 71
  
• Similar to reductions associated with use of
  plain soap
  
• Most Reductions in respiratory illness were not
  statistically significant
  
• 3-20
• Infectious symptoms / absence ranged from 9 to
  43
  
• Variable definitions and reporting

34
Alcohol-based Hand Sanitizer Study Design Issues

• Study Design Limitations
• Unknown etiology for symptom definitions
• Variability in definition of symptoms and
  reporting methods
  
• Not balanced on education intervention
• Lack consistent measurement of alcohol use and
  other hand hygiene practices (soap use?)
  
• Difficult to employ masking
• Only one study controlled for clustering

35
Alcohol-based Hand Sanitizer Conclusions

• Alcohol-based hand sanitizers in conjunction with
  hygiene education can effectively reduce diarrhea
  and general infectious symptoms in the community
  setting
• Alcohol alone?
• There is less evidence of effectiveness for
  reducing upper respiratory infections

36
Future Research Needs

• Assess the effect of antiseptic soaps and
  alcohol-based hand sanitizers in culture
  confirmed viral and bacterial infections studies
  
• Assess the benefit of alcohol-based hand
  sanitizers in groups with similar baseline levels
  of hygiene education
  
• Better control of confounding factors
• Analytical techniques that accommodate clustered
  data
  
• Further household level studies

37
Conflicts of Interest Statement and
Acknowledgements

• Consulted for Pfizer in September
• Thanks to Dr. Elaine L. Larson