Table 3. Data charting vulnerability indices.

Author/year/city country	Title of study	Participants/Data collection	Study Aim	Design	Results: characteristics, medical conditions	Conclusions relating to vulnerability index
Berbesi, Segura, Cardona, Caicedo, (2017) Medellin, Colombia	HIV vulnerability index in homeless persons	N = 338 Data collected first half of 2014, homeless persons on the streets of Medellin (Colombia)	To determine an HIV vulnerability index for homeless persons	Cross sectional study using a HIV vulnerability tool developed by the authors	Age range 18–65 years, 50% of participants were >41 years, 71% (n = 241) single, HIV prevalence was 8.15% (CI 95% 3,92–12,37)	HIV vulnerability is defined as the reduced ability to anticipate (lack of knowledge to protect oneself), resist (risky sexual behaviour and drug use) and recover (lack of social supports), which limits a homeless person’s ability to access HIV prevention and support services
Bowie & Lawson (2018) Seattle, US	Using the Vulnerability Index to Assess the Health Needs of a Homeless Community	N = 46 Data collection over two-months, participants were encountered in drop-in day centres, temporary winter overnight shelter, and on the street in Seattle. All participants met the Federal definition of homelessness	To assess the health status and health service usage of people experiencing homelessness in an urban neighbourhood on the edge of a large city using a vulnerability index	Cross-sectional survey design using interviews to complete the vulnerability index. Interviews conducted by a member of the homeless community paired with a faculty member/ graduate student nurse	Age range 28–66 years, average age 47 years. Majority male (70%), mean continuous duration of homelessness 4.7years, range 1 to 19 years. Common medical conditions: heart disease (37%), skin conditions (28%) and hepatitis C (22%), abuse of drugs or alcohol (78%), mental health condition (44.7%).	Vulnerability index was easy to administer, and effective in compiling a health profile of people experiencing homelessness in the community, enabling workforce planning. Vulnerability index lacks robust psychometric testing so cannot be used for outcome prediction. Vulnerability index lacks a mechanism for interpreting the overall vulnerability score beyond prioritising shelter
Brown, Cummings, Lyons, Carrion, & Watson (2018). Midwest, US	Reliability and Validity of the Vulnerability Index-Service Prioritisation Decision Assistance Tool in real-world implementation	N = 1407 Data retrieved from Homeless Management Information System April 2014-April 2016 VI-SPDAT administered via street outreach	Reliability and validity of the VI-SPDAT	Internal, test-retest and inter-rater reliability and construct and predictive validity of the VI-SPDAT	VI-SPDAT has limitations in its reliability and validity. Test-retest reliability coefficients were below acceptable thresholds. Several items on the Socialisation and Daily Functions and Wellness domains demonstrated negative associations with other variables	Use of the VI-SPDAT in a community context is not recommended as the sole instrument for housing prioritisation, further psychometric testing is required. Total VI-SPDAT score not a predictor in the re-entry to homeless services but correlated with higher risk
Cronley, Petrovich, Spence-Almaguer, & Preble. (2013) Fort Worth, Texas, US	Do Official Hospitalisations Predict Medical Vulnerability among the Homeless? A Postdictive Validity Study of the Vulnerability Index	N = 97 Data from vulnerability index assessment of homeless individuals in 2008. Individuals scoring zero or one on the vulnerability index were not offered housing and not included in the sample. Participants were those who were assessed using the vulnerability index in 2008 and had received health care from the hospital in Tarrant County	Empirical evaluation of the vulnerability index as a tool to assess the degree of medical vulnerability and health service utilisation among people experiencing homelessness	Postdictive validity of the vulnerability index assessment data was paired with the health care utilisation data collected from the Hospital in Tarrant County	Age range 29–69 years mean age 48.69 years, 53.3% African American, 57.7% male. Common medical conditions were heart disease (35.1%), asthma (29.9%), Hepatitis C (23.7%), Tuberculosis (16.5%), Liver disease (16.5%), Emphysema (13.4%). When controlling for gender and race, individuals who scored higher on the vulnerability index accessed hospitals more frequently, compared to those who had a low vulnerability index score	Official hospital records are predictive of overall vulnerability index scores and are correlated with self-reported hospitalisation data but are not predictive of the subcomponents of the vulnerability index, perhaps indicating the underutilisation of health care for those with serious health conditions
Montgomery, Syzmkowiak, Marcus, Howard, Culhane. (2016) 62 communities in the US	Homelessness, Unsheltered Status, and Risk Factors for Mortality: Findings from the 100,000 Homes Campaign	N = 25,489 Total N = 13761 Unsheltered N = 11728 Sheltered Data collected as part of 100,000 Homes Campaign, 2008–2014, in 96 communities	Comparison of characteristics of people experiencing homelessness who were sleeping in unsheltered situations with those who were accessing shelter	Cross sectional survey data collected through the application of the vulnerability index, to assess sheltered status and risk factors for mortality of people experiencing homelessness	Highest proportion of participants were aged between 50–59 years (33.9% vs. 34.3% sheltered vs. unsheltered), majority male (70.2% vs. 75.6% sheltered vs. unsheltered), majority homeless for 1–5 years (47.8% vs. 46.5%, sheltered vs. unsheltered). Common medical conditions, mental health (53.9% vs 53.7%, sheltered vs. unsheltered), living with liver/and or kidney disease (11.8% vs. 15.1%, sheltered vs. unsheltered), HIV/AIDS 3.3% vs. 3.7%, sheltered vs. unsheltered), ever treated for drug/alcohol abuse 45.1% vs 47.2%, sheltered vs. unsheltered)	Unsheltered status correlated with being male, white or mixed race, history of military service, incarceration, foster care, use and treatment for drug and alcohol abuse, less likely to have more than high school education, more likely to receive income from informal sources, higher rates of the high risk conditions measured by the vulnerability index, more likely to live in warmer climates. Sleeping unsheltered had a 12% higher odds of having at least one risk factor for mortality, other correlates for mortality were female, military service, homeless for <5years, prior incarceration. The findings highlight the need to identify those at risk and assist them in their transition from military service or incarceration or foster care, to ensure that they do not become unsheltered and later, at risk of increased mortality
Nicholson, Graham, Emery, Schiff, Giacomin, Tanasescu. (2008) Calgary, Canada	Describing the Health of the Absolutely Homeless Population in Downtown Calgary 2008	N = 132 Data collection October–December 2008, as a street level survey	To describe the health profile of the homeless population in Calgary using the vulnerability index and compare the findings with homeless people in American cities	Survey data collected through the application of the vulnerability index, comparison of findings with homeless people from American cities	Mean age 40 years, 79.5%(n = 105) male, a third identified as Aboriginal (n = 35), average length of homelessness was just under 6 years, 45.5% (n = 60) sleeping most frequently in shelters, 19.6%. 73% reported at least one health condition and 55% had two or more, history of frostbite (25.8%, n = 24), asthma (23.4%, n = 31), hepatitis C (22%, n = 29). 32.6% (n = 43) current or previous treatment for mental health issues, 96.2% (n = 127) reported a substance abuse problem	The vulnerability index tool was effective in assessing the homeless population. Compared with the general population in Canada, participants reported higher incidence of kidney disease, asthma, emphysema and cancer. Females reported higher incidence of cancer, hepatitis C, liver, kidney heart and mental health disease. Homeless people in Calgary have a higher risk of mortality