Guide to using the Healthy Cities Toolkit

About the Toolkit

The Healthy Cities Toolkit aims to understand what factors impact the health and wellbeing of those living and working in cities.

The Toolkit is based on the findings of over 250 systematic reviews. It is designed to support practitioners, policymakers, and the public in improving the health of the population in their cities. 

The Toolkit provides evidence-based summaries of what is likely to benefit or negatively impact health.

The Toolkit pages should be used alongside professional expertise and local knowledge to move from the summarised information to evidence-informed decisions about what might work best in your city context.

Toolkit guide

The Healthy Cities Toolkit is an accessible summary of urban health research and provides a rigorous assessment of over 50 approaches to improving urban health, each summarised in terms of:

  • impact,
  • resource implications, and
  • the quality of the evidence supporting the approach.

What do we mean by ‘impact’?

The impact is the effect that an urban intervention has on individual health or the broader wellbeing of society. The impact was assessed using the reported outcomes in the review. 

The impact is rated as positive or negative. Uncertain impact indicates that the evidence suggests that the intervention or exposure makes no difference, or there is still considerable uncertainty about the overall effect.

ScoreParameters*
Strong positive impact: Statistically significant effect in >75% of studies or pooled meta-analysis
AND
Clinically significant effect on health in >75% of studies or pooled meta-analysis
Moderate positive impact: Statistically significant effect in >50% of studies or pooled meta-analysis
AND 
Clinically significant effect on health in >50% of studies or pooled meta-analysis
Mild positive impact: Statistically significant effect in 50-25% of studies or pooled meta-analysis
OR
Clinically significant effect on health in 50-25% of studies or pooled meta-analysis
Low positive impact: Statistically non-significant effect in >50% of studies or pooled meta-analysis
OR
Non clinically significant effect on health in >50% of studies or pooled meta-analysis
Uncertain: Statistically non-significant effect in <50% of studies or pooled meta-analysis
AND 
Non clinically significant effect on health in <50% of studies or in pooled meta-analysis
Low negative impact: Statistically non-significant negative effect in >50% of studies or pooled meta-analysis
OR 
Non-significant adverse effects on health in >50% of studies or pooled meta-analysis
Mild negative impact: Statistically significant negative effect in 50-25% of studies or pooled meta-analysis
OR 
Significant adverse effects on health in 50-25% of studies or pooled meta-analysis
Moderate negative impact: Statistically significant negative effect in >50% of studies or pooled meta-analysis
AND 
Significant adverse effect on health in >50% of studies or pooled meta-analysis
Strong negative impact: Statistically significant negative effect in >75% of studies or pooled meta-analysis
AND 
Significant adverse effect on health in >75% of studies or pooled meta-analysis

*These parameters are subjective but use quartiles to categorise the included studies and will be reviewed on an ongoing basis. 

What do we mean by ‘resources’?

Resources are the collective requirement for funding, materials, labour, time or other assets required to deliver an urban intervention or the financial impact of the urban exposure on the individual, community, city or organisation. The judgments are subjective and therefore require integrating with the local context, which may affect the resources required.

ScoreParameters *
High resources: >75% of included reviews reported resource requirements, and >50% were ranked as high
Moderate resources: >50% of included reviews reported resource requirements, and >50% were ranked as moderate
Low resources: 50-25% of included reviews reported resource requirements, and >50% were ranked as low
Very low resources: 25% of included reviews reported resource requirements, and >50% were ranked as very low
Uncertain: >75% of included reviews did not report information regarding the cost

*Detailed explanations of the resource implications are summarised on each toolkit page based on the evidence provided in the reviews. These parameters are subjective but use quartiles to categorise the included studies and will be reviewed on an ongoing basis. 

How do we judge the quality of the ‘evidence’?

The quality of the evidence reflects the rigour of research and helps attribute confidence in the results. Well-conducted evidence reviews use “risk of bias” or quality assessments to grade the evidence. 

To assess the quality of the evidence, we used an adapted version of AMSTAR 2, a popular instrument for critically appraising reviews of healthcare interventions and the reported findings from the risk of bias or quality assessments. 

There are seven critical domains of the AMSTAR 2 instrument; we used six of the domains depending on the review methods. For all studies included in our Toolkit, we used the following domains:

  • protocol registration before the commencement of the review (item 2), 
  • adequacy of the literature search (item 4), and 
  • justification for excluding individual studies (item 7). 

For all studies included in our Toolkit, we also extracted the following information: 

  • design of the review,
  • design of the studies included in the review,
  • dates of the search (to assess how up to date the evidence is),
  • numbers and types of databases used to search for evidence, and 
  • other evidence included (e.g. grey literature or backward searching). 

If the included studies did not conduct a risk of bias or quality assessment, then the quality of the evidence was automatically rated as ‘uncertain’. If the included review conducted a risk of bias or quality assessment, then a further two domains of AMSTAR 2 were used, including

  • consideration of the risk of bias when interpreting the results of the review (item 13) and
  • assessment of the presence and likely impact of publication bias (item 15).

If the included studies also conducted a meta-analysis, then item 11 of AMSTAR 2 was used to grade the appropriateness of meta-analytical methods. Item 9 of AMSTAR 2 was not used as this domain was relevant for individual studies rather than reviews.

ScoreParameters*
High quality: >75% of included reviews conducted a risk of bias or quality assessment, and >50% of the assessment scores were ranked as low risk of bias or high-quality
Moderate quality: >50% of included reviews conducted a risk of bias or quality assessment, and >50% of the assessment scores were ranked as moderate risk of bias or quality
Low quality: >50% of included reviews conducted a risk of bias or quality assessment, and >50% of the assessment scores were ranked as high-risk of bias or low quality
Very low quality: >25% of included reviews conducted a risk of bias or quality assessment, and >50% of the assessment scores were ranked as very high-risk of bias or very low quality
Uncertain: <25% of included reviews conducted a risk of bias or quality assessment

*These parameters are subjective, but use quartiles to categorise the included studies and will be reviewed on an ongoing basis. 

Using the Toolkit effectively 

  1. Your professional judgement can be used to apply the evidence: The Toolkit summarises evidence on the impact of approaches tested in the past. In your city, it is crucial to consider the needs of your citizens and any barriers they might face before adopting an approach.
  2. Consider the impact, resources, and quality of the evidence together: Some approaches may be effective but not cost-effective. Other approaches may have a lower overall impact but have a more well-established evidence base.
  3. Read past the headline figure and think about what is ‘behind the average’: For every approach in the Toolkit, the average hides a range of different impacts. The text in the Toolkit describes why effects vary and indicate where approaches may have a greater impact. 
  4. Think about principles of good implementation: Adopting an approach is not a one-off event and requires careful planning and implementation. 

Limitations to the Toolkit 

The Toolkit summarises reviews that have been conducted in the past. It cannot predict how an approach will be received in any city. There are also some aspects where there was insufficient evidence to measure impact, resource implications, and the quality of the evidence flagged as “no impact” or “not clear”. Even with these limitations flagged, it is important to remember that all research in the Toolkit is a description of “what has worked” and not “what works”. By understanding these trends in previous research, the Toolkit will provide cities with information on approaches that are promising in particular contexts and approaches that carry risks.