Skip to main content

Table 6 Champions’ effectiveness on patient, provider and system/facility’s outcomes

From: The effectiveness of champions in implementing innovations in health care: a systematic review

Subcategory (# of studies)

First author, year

Study Design

Champion operationalization

Outcome extracted from included study

Statistical analysis/approach

Test statistic (measure of magnitude)

p-value

Patient Outcomes (n = 6)

Conclusion: Across six studies, there are mixed findings pertaining to use of champions and improvement in patients’ outcomes related to innovation use

Improvement in patient’s health outcomes

(n = 4)

Ben-David, 2019 [51]

Cross-sectional study

Presence of ward infection control champions

(survey item asking if a nurse or/and physician champion was present)

Monthly incidence rates of central-line-associated bloodstream infection

Negative binomial regression

Incidence rate ratio (95% CI) = 0.47 (0.31–0.71)

 < .001

 

Bradley, 2012 [53]

Cross-sectional study

Presence of one or more physician/nurse/ both/no champions

(two “Yes/No” survey items asking the presence of physician/nurse champions)

30 days risk-standardized mortality rate post acute myocardial infarction in hospitals

Multivariate linear regression

β (95% CI) =  − 0.695 (− 1.253 to − 0.137)

(No champion vs nurse champion only)a

.015

β (95% CI) =  − 0.731 (− 1.404 to − 0.059)

(Physician champions vs nurse champion only)

.033

β (95% CI) =  − 0.880 (− 1.442 to − 0.318)

(Both physician and nurse champions vs nurse champion only)

.002

 

Ward, 2004 [76]

Cross-sectional study

Presence of champion

(single item rated on a 5-point Likert scale)

Patient outcome measures relative to guideline-based diabetes management

Single predictor generalized estimating equations

β (SE) =  − 0.38 (0.39)

.3202

 

Whitebird, 2014 [41]

Prospective cohort

Presence of a strong primary care provider champion

(“Yes/No” extracted from quality improvement narrative reports)

Average monthly remission rates at 6 months (number of patients with a score of < 5 on the PHQ-9)

Pearson's correlation coefficient

r (95 CI) = 0.40 (− 0.16 to 0.77)

 > .05

Quality of life (n = 1)

Hung, 2008 [60]

Cross-sectional study

Presence of practice (health promotion) champions (single item rated on a 5-point Likert scale)

Fewer numbers of unhealthy days in the past 30 days

Hierarchal generalized linear modelling

β (SE): 0.34 (0.07)

OR (95 CI) = 1.41 (1.22–1.64)

 < .001

  

Fewer numbers of limiting days in the past 30 days

 

β (SE): 0.53 (0.19)

OR (95 CI) = 1.71 (1.16–2.53)

 < .01

    

General health status

 

β (SE): 0.38 (0.09)

OR (95 CI) = 1.47 (1.20–1.79)

 < .001

Patient Experience

(n = 1)

Goff, 2019 [57]

Cross-sectional study

Presence of a designated quality champion

(“Yes/No” survey item)

Average patient experience scores of clinics that are part of the Massachusetts Health Quality Partners (MHQP)

ANOVA

Test statistics not reported (Mean difference = 0.09 favouring presence of a champion)

.29

Provider Outcomes (n = 1)

Conclusion: There is a single study suggesting that the use champions is related to improvements in provider outcomes related to innovation use

Satisfaction with practice (n = 1)

One study, two reports:

Weiler, 2012, 2013 [77, 78]

Interrupted time series

Endorsed by champions (three items rated at a 6-point Likert scale based on Mullins et al. [98]

Reported ergonomic advantage 1-month post-introduction of transfer boards

Structural equation model using a maximum likelihood method

Path coefficients (95 CI) = 0.63 (.0664–1.1936)

 < 0.05b

System/Facility Outcomes (n = 3)

Conclusion: Across three studies, there is a trend suggesting that use of champions is related to improvement in system/facility outcomes related to innovation use

Hospital quality of care indicators

(n = 1)

Foster, 2017 [56]

Non-controlled before and after study

An average of 0.1 champion fellows in 1160 hospitals (number of champion fellows)

Weighted composite score of quality of care—occurrence of 10 harm topics (e.g. readmissions) for 1 month

Multivariate linear regression

Adjusted effect over time: β =  − 0.9 (negative β = more effective in this study)

.008

Access to training for alcohol cessation screening and advice

(n = 1)

Patton, 2013 [69]

Cross-sectional study

Presence of champion (“Yes/No” survey item)

Emergency staff’s access to training for screening for alcohol consumption

Chi square

χ2 = 36.64

 < 0.001

    

Emergency staff’s access to training for providing brief advice regarding alcohol consumption

 

χ2 = 29.93

 < 0.001

Compatibility and sustainability of in-house pharmacy immunization services

(n = 1)

Westrick, 2009 [79]

Cross-sectional study

Strategic champion effectiveness

(4-item scale on champion’s commitment, advocacy and ability to manage and acquire resources) adapted from Hays et al. [99]

Compatibility between immunization services and host pharmacy

Multivariable linear regression

β = 0.12

.300

    

Sustainability of in-house pharmacy immunization services

 

β = 0.00

.978

   

Operational champion effectiveness (4-item scale on champion’s knowledge, ability to manage an in-house immunization service, and to resolve conflicts) adapted from Hays et al. [99]

Compatibility between immunization services and host pharmacy

Multivariable linear regression

β = 0.31

.005

    

Sustainability of in-house pharmacy immunization services

 

β = 0.09

.419

Other Outcomes (n = 1)c

Adaptation and evaluation of in-house pharmacy immunization services

(n = 1)

Westrick, 2009 [79]

Cross-sectional study

Strategic champion effectiveness

(4-item scale on champion’s commitment, advocacy and ability to manage and acquire resources) adapted from Hays et al. [99]

Degree of modifications made to in-house pharmacy immunization services

Multivariable linear regression

β = 0.05

.705

    

Formal evaluation of in-house pharmacy immunization services

 

β = 0.26

.038

   

Operational champion effectiveness (4-item scale on champion’s knowledge, ability to manage an in-house immunization service, and to resolve conflicts) adapted from Hays et al. [99]

Degree of modifications made to in-house pharmacy immunization services

Multivariable linear regression

β = 0.05

.698

    

Formal evaluation of in-house pharmacy immunization services

 

β = 0.09

.419

  1. aIn this study, groups exposed to only nurse champions had the highest risk-standardized mortality rate (RSMR; RSMR = 16.2); hence, it was the reference variable
  2. bThe authors reported a path coefficient that they stated is significant at a p-value of 0.1. Manual calculation of the 95% CI was done by JES to determine significance of both ergonomic advantage and intention to use at a p-value of .05
  3. cOther outcomes were not considered in analysis