Social resources during COVID-19

By Jillian Macklin, MD/PhD candidate at the University of Toronto & Rachelle Perron, RN and eMHI candidate at the University of Toronto

What did we do?

The Upstream Lab worked with the Centre for Effective Practice and the Department of Family and Community Medicine at the University of Toronto to develop the COVID-19 Social Care Guidance. This helps providers support people during COVID-19. It provides a suggested script and questions to ask each person about key social determinants of health: income, food, housing, social connection and protection from violence. This Guidance also includes a framework and approach for engaging in conversations with people grounded in empathy and non-judgemental care.

COVID-19 Social Care Guidance

When providers pose these questions we want to ensure they have resources at their fingertips. We worked with over 60 volunteers across Canada to compile lists of social resources across Canada. We learned a lot about rapid collaboration and managing a team, and are so thankful for our volunteers’ efforts!


Why does this matter?

The COVID-19 pandemic is having a disproportionate impact on marginalized populations. Public health measures aimed at limiting the spread of COVID-19, such as physical distancing, self-isolation and the closure of non-essential businesses, have the potential to have the unintended consequence of worsening social conditions that can lead to poor health. Early evidence suggests that food insecurity, social isolation, housing insecurity, unemployment and interpersonal violence have increased. People made vulnerable by how our society is organized have found themselves with even less access to the resources and supports needed to keep healthy and safe.

Health and social care providers come in frequent contact with people who are struggling with complex social issues and are well suited to assist these individuals to connect with services that address social needs. However, since COVID-19 began, many community organizations have closed or altered their service delivery and numerous new resources and programs have been created. For health and social care providers, a central problem is being able to quickly connect individuals with the right resources at the point-of-care.

What are our next steps?

Our next step is an online COVID Social Resource Connector that will help providers across Canada link clients with social supports during COVID-19. The Connector will help providers screen for unmet social needs, identify relevant and appropriate supports, and complete referrals to community resources. Through our site, community organizations will be able to share service information and accept referrals, free of charge. By supporting health and social care providers to address the social determinants of health during COVID-19, we hope that this Connector will help reduce the social harms of the COVID-19 pandemic.

Collecting data on race during COVID-19: Recommendations

Last week, evidence emerged from the US that COVID-19 outcomes may be associated with race. Our COVID-19 page provides more detail. Briefly, numerous jurisdictions reported that Black individuals were significantly over-represented in COVID-19 related deaths. Some very early data has been brought together by the CDC.

This led to many people asking: Are the same health inequities happening in Canada? We unfortunately do not routinely collect this data in health care settings. Shree Paradkar explored this issue in an article for the Toronto Star, and I had a chance to share some of our experiences with collecting data on race in primary care, and our efforts with the SPARK Study. The issue has become more contentious as public health leaders have put forward different views on whether such data should be collected.

We have put together a short report on this issue, and include recommendations. We hope this is useful during COVID-19 and beyond.

Community Health Workers to address unmet social needs: RCT

Our Upstream Lab journal club focused on this interesting study: Effect of community health worker support on clinical outcomes of low-income patients across primary care facilities: a randomized clinical trial (JAMA Intern Med 2018; 178(12): 1635-1643). This is led by Dr. Shreya Kangovi, who is the founding director of the Penn Centre for Community Health Workers.

At the heart of this study is the intervention called IMPaCT, which the Penn Centre website describes as “a standardized, scalable program that leverages community health workers (CHW) –trusted laypeople from local communities– to improve health.” In the paper, the authors note 3 stages:

  1. Goal-setting using a semi-structured interview guide to collect data on social determinants, open-ended questions and asking participants what they need to reach their health goals. Strategies discussed are familiar to those who motivational interviewing (MI) techniques in their clinical care.
  2. Tailored support over 6 months, including weekly contact and meeting face-to-face at least once a month.
  3. Connection with long-term support, which could include neighbours, family or even a weekly CHW facilitated support group.

Getting to the results, in this RCT comparing goal setting alone (n=288) vs goal setting plus CHW (n=304), there was no difference in self-rated health at 6 months (assessed using the SF-12v2 Physical Component). However, those in the intervention group had a substantial reduction in total days in hospital (absolute event reduction of 69% at 6 months), due to shorter average length of stay and lower mean number of hospitalizations.

This is quite striking, and in sharp contrast to our study from last week, which found no impact on health service use. Dr. Kangovi and colleagues recently published a study in Health Affairs finding a positive return on investment of IMPaCT, where every dollar invested in this intervention would return $2.47 to the average payer (e.g. Medicaid) within the fiscal year.

We’ll certainly apply lessons learned to our work, particularly the SPARK Study, but also several other projects where we screen and intervene on social needs at the individual-level.

Negative RCT of health care hotspotting in NEJM: What outcomes are important?

At the Upstream Lab, we’ve started up a regular journal club as part of our weekly meetings. We’ll be sharing a brief summary here on our blog. This week we looked at this study: Health care hotspotting – a randomized controlled trial (NEJM 2020; 382: 152-162). This was also the focus of discussion last week at MAP, the centre where we are based.

You may remember Atul Gawande‘s 2011 article on “The hot spotters”, profiling the work of the Camden Coalition of Healthcare Providers in Camden, New Jersey. This article got a lot of people interested in the idea that the solution to rising health costs was to focus on the people who were disproportionately using a huge amount of health services. Here in Ontario, we also started to focus on so-called “high-cost users”, leading to a new program called Health Links. I was involved in a study led by my colleague Dr. Laura Rosella that essentially found that social determinants played a big role in who would end up being a high-cost user. Maybe not surprising to visitors of this blog, but we thought it was really important to point this out. Most of the solutions being proposed were focused on just getting these folks more traditional health services.

Getting back to the article, this RCT focused on “superutilizers of the health care system — persons with medically and socially complex needs who have frequent hospital admissions”. These patients were identified during an admission and were connected to a multidisciplinary team with nurses, social workers, community health workers and health coaches. This team did a lot, including home visits, got people to appointments, helped with medications, measured BP and blood sugars, coached patients and helped patients apply for social services and behavioral health programs. That last part about social services sounds similar to some interventions we are studying.

The study found no difference in readmission to hospital after 6 months, between people who got the intervention compared to controls. Why not? Regression to the mean, is one explanation . Another is that people are on trajectories that are hard to change. A NYTimes article on the study made the case that it is really tough to make a difference in readmission to hospital, and mentions future work by the Camden Coalition will focus on housing. They are teaming up with lawyers to prevent evictions, something we’ve also been studying with the Health Justice Initiative.

For such studies, what outcomes are important to patients, families, providers and policymakers? I suspect that if outcomes included reduced stress, improved quality of life, and knowing where to go for help when in crisis, maybe the study would have been positive. Like so many studies of complex interventions, more work is required to really understand what is going on.

Tools for upstream thinkers: 5 tips for effective policy briefs

madeleine-bondy.jpgby Madeleine Bondy, 2019 summer student at the Upstream Lab

Too often researchers are not reaching out to those in government. As a young researcher and future physician, I’m curious about how to best influence the policies that impact the social determinants of health. Translating research into action is a key part of our work at the Upstream Lab.

Policy briefs are documents used to inform political action and influence public policies. They are usually short 3-4-page documents, but can be as longer. The briefs summarize an issue and include a synthesis of research evidence with the purpose of orienting policy action.

In writing a brief it is important to include the purpose of the brief as well as the problem and current status, the context, key considerations and existing evidence, an overview of the solutions including pros and cons and then your policy recommendation.

5 tips to make policy briefs effective

1. Know your audience

Who  the brief for and how will they use? Consider your audience’s needs and expectations. What is on their mind, what are the questions they need answered today? You may need different versions of a brief for different audiences. For example, one written for a specialist audience may need to be longer (6 pages) than for one for non-specialists (2-4 pages).

2. Context matters

Consider how the policy brief fits into the overall context of your knowledge and sharing strategy. A policy brief may not always be the most appropriate tool for your purposes. Rather, a phone or in-person chat may be best, followed by a brief. Avoid viewing the brief as an end to itself, but rather part of a wider strategy.

3. Power and interest are paramount

Policy is formed at multiple levels, so take time to select the right group to share your research. Your organization may not be the most appropriate one to disseminate the results so you may need to build relationships with other organizations that are better positioned to reach your intended audience. To reach the general public, for example, it is usually best to engage the media, using social media and online platforms. The graphic below shows how it is important to think about organizations that have the power to either support or oppose the issue, and may have potential interest in your subject.

4. Policy briefs are a dialogue

A policy brief doesn’t contain every detail about a topic, but sparks a dialogue. The most effective briefs draw attention to a specific problem, provide a clear interpretation of the results and describe a limited number of actions to be taken. The greater amount of trust in the authors of the policy brief, the higher the likelihood that its recommendations will be well received.

5. Understand the potential and the limitations of briefs

Researchers should produce policy briefs but should also be realistic about the outcomes. Policy briefs are only one part of your knowledge-sharing strategy. Continue to look for policy windows to re-disseminate the brief (i.e. when political circumstances change).

Short, medium-length and long examples of policy briefs focused on the social determinants of health:

  • The Wellesley Institute published this 2-page brief in 2017 on the need for supportive housing in Ontario.
  • The Hispanic Health Council published this 22-page brief on how community health workers can help address social determinants
  • Texas Children’s Hospital and Baylor College of Medicine published this 52-page brief on screening for social determinants in clinical settings.


Based in part on a webinar on how policy briefs can help bridge the research-to-policy gap, presented on June 19, 2019 by the National Collaborating Centre for Public Policy. Recording available here.