Self-reported oral health of a metropolitan homeless population in Australia: comparisons with population-level data


Dr Eleanor J Parker
School of Dentistry
Faculty of Health Sciences
The University of Adelaide
Adelaide SA 5005


Background:  There is limited information on self-perceived oral health of homeless populations. This study quantified self-reported oral health among a metropolitan homeless adult population and compared against a representative sample of the metropolitan adult population obtained from the National Survey of Adult Oral Health.

Methods:  A total of 248 homeless participants (age range 17–78 years, 79% male) completed a self-report questionnaire. Data for an age-matched, representative sample of metropolitan-dwelling adults were obtained from Australia’s second National Survey of Adult Oral Health. Percentage responses and 95% confidence intervals were calculated, with non-overlapping 95% confidence intervals used to identify statistically significant differences between the two groups.

Results:  Homeless adults reported poorer oral health than their age-matched general population counterparts. Twice as many homeless adults reported visiting a dentist more than a year ago and that their usual reason for dental attendance was for a dental problem. The proportion of homeless adults with a perceived need for fillings or extractions was also twice that of their age-matched general population counterparts. Three times as many homeless adults rated their oral health as ‘fair’ or ‘poor’.

Conclusions:  A significantly greater proportion of homeless adults in an Australian metropolitan location reported poorer oral health compared with the general metropolitan adult population.


In 2006, 105 000 Australians were considered homeless: a rate of 53 persons per 10 000 of the population.1 In South Australia, the City of Adelaide had 1.5% of the broader metropolitan Adelaide’s total population, but 15% of the homeless population, numbering 762 homeless people.2 It is common for the inner city to have higher rates of homelessness as people tend to cluster where services are located.2

There are multiple definitions of homelessness, which can vary both within and between nations, making direct comparisons difficult. The Australian Bureau of Statistics utilizes a definition that is comprehensive in its inclusion of accommodation states, but also separates homelessness into subgroups of ‘primary’, ‘secondary’ and ‘tertiary’ homelessness.1‘Primary’ homelessness includes those without conventional accommodation, such as sleeping in parks or streets or improvised shelters.1‘Secondary’ homelessness includes those in emergency accommodation, supported accommodation programmes, residing with others because they have no accommodation and staying in boarding houses. ‘Tertiary’ homelessness includes those living in boarding houses for 13 weeks or more and whose accommodation is below the minimum community standard.1

Indigenous Australians are over-represented in the homeless population, being 2.4% of the national population but 9% of the homeless population.1 Men are also over-represented. In South Australia, amongst the total homeless population, men outnumbered women, 54% to 46%, but from age 35 onwards, men outnumbered women 63% to 37%.2

Reviews of the literature have demonstrated high levels of general health problems amongst homeless populations internationally, with homelessness being clearly associated with poor health.3,4 Homeless people are a unique population in which substance abuse and mental health issues frequently feature.2,5 A minority of people have both substance and mental health problems.2 There are a limited number of studies reporting on the oral health of homeless populations and, in particular, their self-reported oral health. The published studies, although tending to focus on specific geographical groups, have consistently demonstrated poor oral health and high oral health care needs amongst homeless groups. This has included higher numbers of carious and missing teeth than the general population, high levels of oral health impact, high levels of perceived need but low service utilization and understanding of how to access care, as well as poor self-reported oral health.6–11 Discrepancies in the perceptions of service needs between clients and care providers have been demonstrated, with homeless clients with mental illness identifying the need for dental care more frequently than support workers.12 This has potential implications as homeless people often rely on support centres to advocate for and facilitate access to services.

Although access to services among socially disadvantaged groups is recognized, homeless populations are an especially underprivileged group, usually excluded in population-level surveys because the very nature of their existence means they are not contactable through traditional recruitment measures (e.g. telephone or postal surveys). For this reason, it is difficult to ascertain the evidence-based oral health service needs of this population subgroup.

It is important that the oral health needs of this severely disadvantaged group are quantified, and perceptions of homeless people explored, so that appropriate policies and public health strategies that are context-specific can be specifically implemented among such groups. This is relevant across all nations.

The aim of this study was to quantify and compare self-reported oral health among homeless adults in Adelaide City, South Australia, with data from a representative age-matched population in metropolitan South Australia.


Survey of adults experiencing homelessness

This was a cross-sectional study of a convenience sample of adults utilizing services for homeless groups in the Adelaide central business district. Administration of a self-report questionnaire occurred during a three-week period in 2009.


Participants were recruited via attendance of questionnaire administrators at a total of five accommodation facilities and drop-in centres where low-cost meals and a range of support services are provided. Sessions were advertised at each centre through the use of flyers, posters and through centre staff. Any person present when questionnaire administrators were in attendance was invited to complete a questionnaire. Group announcements of an open invitation to complete a questionnaire were made, with centre staff promoting the study. A total of nine questionnaire administrators followed the same protocol for recruitment and questionnaire administration, with each session supervised by one of the study authors. Each centre was visited two or three times during the three-week period. Due to the nature of recruitment, with minimal direct one-on-one recruiting, it was not possible to record the number of non-consents.


Participants needed to be aged 17+  years, residing in an accommodation facility or attending a ‘drop-in centre’ for homeless people, and be able to understand and communicate in spoken English. Participants were deemed fit to participate if they were able to coherently communicate what they were consenting to. If there were concerns about comprehension or consent, questionnaire administration did not proceed.


Participants received a $20 supermarket voucher upon completion of the survey, as well as a toothbrush, toothpaste and dental floss.

Ethical approval

Ethics approval was granted by the Human Research Ethics Committee of the University of Adelaide and the Aboriginal Health Council of South Australia. Participants gave signed informed consent before participating. If participants had limited reading ability, the consent form was read to them.

Self-reported questionnaire

The questionnaire was developed specifically for this study, based on selected items used by the Australian Research Centre for Population Oral Health in other population-level surveys and surveys of disadvantaged groups. Questionnaires were either completed as an interview or were self-completed. The level of self-completion was determined by participants, with all questionnaires being reviewed by the interviewer to ensure completion. The questionnaire took approximately 10 minutes to complete.

Population data: National Survey of Adult Oral Health

Population data are from the National Survey of Adult Oral Health (NSAOH) 2004–2006,13 a cross-sectional study of oral health among Australians aged 15 years or more living in all states and territories.


NSAOH utilized a three-stage, stratified clustered sampling design, with the target population being the Australian adult population. The first stage selected postcodes, the second stage selected households within sampled postcodes and the third stage selected one adult from each sampled household.

Ethical approval

Ethical approval was received by the University of Adelaide’s Human Research Ethics Committee. Participants provided verbal consent prior to answering questions in the telephone interview.

Data retrieval for comparison

For the purposes of this study, only data for age-matched participants living in metropolitan Adelaide were used for the items that were comparable. Data were weighted to represent population-level estimates.

Outcome variables

Six domains were considered as outcome variables: (1) demographics; (2) social determinants of oral health and health behaviours; (3) dentate status; (4) dental service utilization; (5) self-perceived need; and (6) self-rated oral health.


Demographic information included age, gender, Indigenous status and whether or not participants were Australian-born.

Social determinants and health behaviours

Information pertaining to oral health related social determinants included highest educational qualification, income source and government health care card ownership, and for health behaviours, tobacco smoking status.

Dentate status

Items pertaining to dentate status included any remaining natural teeth, previous extractions and number of teeth extracted.

Dental service utilization

Dental service utilization items included prior visiting, duration since last visit, place of last visit, usual reason for visiting, avoidance of dental care due to cost and difficulty paying a $100 dental bill.

Self-perceived need

Self-perceived need was quantified by asking participants their perceived need for fillings or extractions.

Self-rated oral health

Participants were requested to rate their general and oral health. Responses provided on a four- or five-point Likert scale were dichotomized for analysis purposes.

Data analysis

For each item, the prevalence and 95% confidence intervals (95% CI) were determined. Where the corresponding 95% CI for the two groups were not overlapping, statistical significance was assumed.13


Completed questionnaires were obtained from 248 participants (age range 17–78 years), 79% of whom were male. There were a greater proportion of individuals less than 40 years of age, more males, more Indigenous and more Australian-born in the homeless sample in comparison to the age-matched metropolitan NSAOH data (Table 1).

Table 1.   Demographics of homeless adults (n = 248) with comparison to age-matched metropolitan population data (n = 324)
 Homeless % (95% CI)NSAOH SA Metro % (95% CI)
  1. *Non-overlapping confidence intervals, indicating statistical significance.

 17–40 years50.8 (42.0–59.6)37.0 (29.5–45.1)*
 41+ years49.2 (40.3–58.1)63.0 (54.9–70.5)
 Male79.0 (73.3–84.7)45.5 (39.9–51.2)*
 Female21.0 (9.9–32.1)54.5 (48.8–60.1)
Indigenous status
 Indigenous 27.8 (17.2–38.4)0.9 (0.2–3.6)*
 Non-Indigenous72.2 (65.6–78.8)99.1 (96.4–99.8)
Born in Australia
 Yes85.5 (80.7–90.3)73.6 (66.4–79.8)*
 No14.5 (2.9–26.1)26.4 (20.2–33.6)
Current residence
 On the streets or emergency accommodation37.0 (27.0–47.0) 
 Supported residence facilities16.3 (4.8–27.8) 
 Common ground11.4 (−0.4–23.2) 
 Other (friends’ couches, family, camp-grounds)35.4 (25.3–45.5) 

The proportion of homeless people reporting their highest qualification as primary school or high school was twice that of their age-matched counterparts (Table 2). A greater proportion of homeless adults were on a government concession card (Centrelink), with three times as many homeless adults owning a government health care card. Twice as many homeless adults reported previous or current tobacco use.

Table 2.   Social determinants, health behaviours and health of homeless adults (n = 248) with comparison to age-matched metropolitan population data (n = 324)
 Homeless % (95% CI)NSAOH SA Metro % (95% CI)
  1. *Non-overlapping 95% confidence intervals, indicating statistical significance.

Highest qualification
 Primary school15.4 (3.9–26.9)7.8 (5.2–11.5)*
 High school58.9 (50.9–66.9)28.7 (22.2–36.2)
 Trade or TAFE21.5 (10.4–32.6)32.4 (24.2–41.9)
 University4.1 (−8.3–16.5)31.2 (23.7–39.8)
 Job4.8 (−7.4–17.0)65.2 (58.3–71.5)*
 Centrelink95.2 (92.5–97.9)34.8 (28.5–41.7)
Health care card ownership
 Yes89.9 (85.9–93.9)29.9 (23.0–37.9)*
 No10.1 (−1.8–22.0)70.1 (62.1–77.0)
 Currently smoke or used to88.7 (84.5–92.9)47.3 (38.2–56.6)*
 Have never smoked11.3 (−0.5–23.1)52.7 (43.4–61.8)

A greater proportion of homeless adults reported previous dental extractions than their age-matched NSAOH counterparts (Table 3). Twice as many homeless adults reported last visiting a dentist more than one year ago and over three times the proportion of homeless adults reported last visiting a government-funded ‘public’ dental clinic. Twice the number of homeless adults reported their usual reason for a dental visit being for a problem. The proportion of participants reporting avoiding dental care due to cost was greater among the homeless participants, with over twice the proportion of homeless adults reporting ‘a little’ or ‘a lot’ of difficulty paying a $100 dental bill compared with their NSAOH age-matched counterparts.

Table 3.   Dentate status and dental service utilization of homeless adults (n = 248) with comparison to age-matched metropolitan population data (n = 324)
 Homeless % (95% CI)NSAOH SA Metro % (95% CI)
  1. *Non-overlapping 95% confidence intervals, indicating statistical significance.

Any teeth left
 Yes88.7 (84.5–92.9)100.0*
 No11.3 (−0.5–23.1)0.0
If yes, ever had any teeth extracted
 Yes79.8 (74.0–85.6)61.7 (53.8–68.7)*
 No20.2 (8.7–31.7)38.3 (31.1–46.2)
Number of teeth extracted
 Four or less43.9 (32.9–54.9)35.5 (30.0–41.4)
 Five or more56.1 (46.4–65.8)64.5 (58.6–70.0)
Seen a dentist before
 Yes95.6 (93.0–98.2) 
 No4.4 (−7.8–16.6) 
If yes, when did last see a dentist
  Less than one year ago26.1 (15.3–36.9)63.4 (56.2–70.1)*
  More than one year ago73.9 (64.3–83.5)36.6 (29.9–43.8)
If yes, where did last see a dentist
  Public facilities74.3 (67.9–80.7)21.7 (16.3–28.3)*
  Private25.7 (14.9–36.5)78.3 (71.7–83.7)
If yes, what is usual reason for seeing dentist
  Problem84.5 (79.5–89.5)41.5 (33.6–49.9)*
  Check-up15.5 (3.9–27.1)58.5 (50.1–66.4)
During the last year, have you delayed or avoided dental care because of cost
  Yes61.3 (53.5–69.1)35.6 (28.1–44.0)*
  No38.7 (28.9–48.5)64.4 (56.0–71.9)
How much difficulty would you have paying a $100 dental bill
  None or hardly any12.5 (0.8–24.2)58.2 (49.4–66.4)*
  A little or a lot87.5 (83.1–91.9)41.8 (33.6–50.6)

The percentage of homeless adults reporting a perceived need for fillings or extractions was more than twice that of their age-matched counterparts (Table 4). More homeless adults reported ‘fair’ or ‘poor’ general health and three times the proportion of homeless adults reported ‘fair’ or ‘poor’ oral health when compared with their age-matched NSAOH counterparts.

Table 4.   Self-perceived need and self-rated general and oral health of homeless adults (n = 248) with comparison to age-matched metropolitan population data (n = 324)
 Homeless % (95% CI)NSAOH SA Metro % (95% CI)
  1. *Non-overlapping 95% confidence intervals, indicating statistical significance.

Do you think you need to have fillings or extractions
 Yes67.8 (60.7–74.9)31.8 (25.9–38.4)*
 No32.2 (21.8–42.6)68.2 (61.6–74.1)
Would you rate your general health as:
 Excellent, very good or good64.9 (57.5–72.3)87.6 (81.5–91.8)*
 Fair or poor35.1 (25.0–45.2)12.4 (8.2–18.5)
Would you rate your oral health as:
 Excellent, very good or good39.1 (29.3–48.9)79.9 (73.6–85.0)*
 Fair or poor60.9 (53.1–68.7)20.1 (15.0–26.4)


This study aimed to quantify and compare self-reported oral health among homeless adults in Adelaide City, South Australia, with data from a representative age-matched population in metropolitan South Australia. A higher proportion of homeless adults in an Australian metropolitan location reported poor oral health than the general metropolitan adult population.

Before discussing the findings in more detail, it is important to acknowledge the study’s limitations. This was a convenience sample, meaning that the findings cannot be generalized to the broader homeless population in Adelaide. However, the nature of this population’s existence means that standard representative sampling approaches are not appropriate. The methods used were consistent with those implemented in other studies among homeless populations, relying on a convenience sample, snow-balling technique and recruiting participants through programme and support workers.7,8,10,12

Another limitation of the study was the non-specific application of the classifications of ‘homelessness’. Due to the difficulty in accessing this marginalized group, all people attending drop-in centres and accommodation venues designed to cater to the needs of homeless people were assumed to classify as experiencing either primary, secondary or tertiary homelessness, having previously experienced chronic homelessness or to be at risk of homelessness and relying on subsidized food and support services.2 All participants were included in the analysis as it was deemed difficult and a potential barrier to developing a relationship with this group, to question in detail their living circumstances.

The current study utilized a combination of self-complete and interview techniques. Other researchers have indicated that verbal administration of questionnaires is appropriate for this population subgroup due to low literacy and difficulty in comprehension and concentration.7,8 It was deemed appropriate in this study to leave the decision to participants. However, at completion, all questionnaires were reviewed by an administrator with the participant, to ensure that any ambiguity regarding item content was alleviated.

Shortcomings aside, this was a reasonably-sized self-report investigation for such a marginalized group for a topic for which little is known and even less understood. The findings may be helpful in highlighting areas where future oral health related research and policy development could be directed, e.g. high self-reported need for oral care and the high impact of cost on accessing dental services.

The finding that only one-quarter of the sample visited a dentist in the previous year was not surprising considering the financial barriers identified, with nearly two-thirds identifying avoidance due to cost. Consistent with this are the findings of Conte and colleagues,7 that only 28% of their sample reported visiting a dentist in the previous year and, alarmingly, that one-third did not know where to seek care if they needed it. In a study of homeless adults in Stockholm, over 90% of participants reported avoiding dental care due to economic reasons.6 In Adelaide, the avoidance of dental care due to cost was found despite those with Centrelink benefits (95% of this sample) being eligible for subsidized care through government dental clinics. This highlights the perceived burden of co-payments. This also highlights other potential barriers such as waiting lists, perceived discrimination and not having a fixed address to receive appointments. The only barrier to care explored in this study that was comparable to NSAOH was the financial component. It is important that future work explores other barriers to care to enable programmes aimed at serving this needy population to address them.

The high rates of smoking amongst this homeless population are consistent with other studies.10,11,14 Collins and Freeman10 highlighted that the homeless population in their study were at greater risk of oral cancer, with smoking being a contributing factor. Although not investigated in this study, tobacco smoking is one of the strongest predictors of periodontal disease, an important condition that may influence other chronic conditions such as diabetes and cardiovascular disease.15

In the current study, over two-thirds reported a perceived need for fillings or extractions. This is consistent with the findings of Luo and McGrath,8 where 70% of participants indicated a need to see a dentist. Perceived need for care has been low in other studies16 and previously in Australia, nearly 60% of homeless men reported that they were happy with the state of their teeth and mouth.17 Interestingly, discrepancies in the perceptions of service needs between client and provider have been demonstrated, with a significantly greater proportion of clients reporting a need for dental services than their mental health agency support workers.11 In that study, nearly three-quarters of clients who were homeless and had a mental illness identified a need for dental care, with 15% rating dental services among their three most important needs.12 This indicates that oral health may be a high priority for homeless adults, but other barriers to accessing care may prevent appropriate utilization of available services.

For the homeless participants in the current study, there was a significant difference in the ranking of oral health and general health, which was not demonstrated in the general population sample, with a significantly greater proportion of homeless adults rating their oral health as ‘fair’ or ‘poor’ than did their general health. This suggests that for this population group, oral health is viewed quite separately to general health with regard to impact: even if general health is perceived as good, oral health can still affect quality of life. The high proportion of homeless individuals giving a low ranking to their oral health is not surprising, given international findings.8,11,14 Luo and McGrath8 reported that 58% of participants ranked general health as less than good and 89% ranked oral health status as less than good, with over half rating oral health as ‘poor’ or ‘very poor’. Gibson and colleagues14 reported that over two-thirds of homeless veterans in their study rated their oral health as ‘fair’ or ‘poor’.

In a study of a Belfast homeless population, the authors concluded that the health and psychosocial factors associated with homelessness needed to be incorporated into context-sensitive oral health care delivery.10 Gibson and colleagues reported that dental treatment improved self-rated oral health among homeless veterans in the United States.14 Hyde and colleagues reported that a dental intervention improved employment and quality of life among a group of welfare recipients.18 This suggests that dental care not only improves feelings of self-worth, aesthetics and eliminates pain, but improves the likelihood of being employed which may lead, in turn, to a sustainable change in living circumstances. Given the high rates of poor self-rated oral health in our sample and the avoidance of dental care due to cost, it suggests the need for programmes to address these issues to enable oral health care to contribute to improved overall health and well-being for this population.


The findings of this study clearly demonstrate poor self-reported oral health among homeless adults in a South Australian metropolitan setting relative to their age-matched general population counterparts. This disadvantage is expected to exacerbate the challenges that this vulnerable population face. The Australian National Oral Health Plan19 identifies oral health care for people with low income and social disadvantage, including people who are homeless, as a priority action area. The findings of the current study provide further reason for the development of comprehensive, appropriate programmes addressing the oral health needs of homeless people in Adelaide, South Australia, but also in other settings throughout the world where homelessness is present, to be considered a high priority. It is anticipated that incorporating oral health promotion, preventive and treatment programmes within broader health programmes designed specifically for this marginalized population, will have the most impact. It is important to share and learn from successful strategies internationally that address the needs of homeless people.


Thank you to the volunteer students who assisted in administering the questionnaire: Y Kho, S Verma and J Sharpe. This study would not have been possible without the enthusiasm and support of staff at each of the centres visited. Thank you to Bev Ellis for managing the data entry. Finally, thank you to all of the participants who were willing to complete the questionnaires and share their personal experiences with us. This project was supported by an Australian Dental Research Foundation grant.