Substance use by homeless clinic attenders in Sydney

Abstract

Objective

Describe patterns of substance use and comorbid conditions among clinic attenders in homeless shelters in Sydney.

Method

Retrospective cohort study of 2498 people who attended a psychiatric clinic at one of three homeless hostels between February 2008 and May 2020. Multivariable logistic regression was used to identify factors associated with self-reported substance use, psychiatric diagnosis and measures of social function.

Results

A total of 2041 of the 2498 (81.7%) reported the harmful use of at least one substance, with alcohol (61.8%), cannabis (50.9%) and stimulant drugs (34.9%) the three most common. Those reporting the regular use of two or more substances (1466, 58.7%) were more likely to have a history of early life and adult trauma, a diagnosis of personality disorder, a criminal conviction, receive the Disability Support Pension, be chronically homeless and sleep in the open.

Conclusions

The study found a high rate of polysubstance use among homeless clinic attenders in Sydney, and an increase in the prevalence of substance use compared to previous studies. Substance use is both a cause and a consequence of homelessness, and services to address substance use have to be part of any program to reduce homelessness and sleeping in the open.

Keywords

homelessness substance use mental health prevalence trauma

There is a high prevalence of substance use disorder (SUD) among homeless populations in high income countries¹ and rates of substance use among homeless people in Australia are reported to be higher than elsewhere.^2,3 Estimates of the prevalence of alcohol use disorder in homeless populations range from 8.1% to 58.5%, and the range from 4.5% to 54.2% for other drug dependence, influenced by both age and gender and the relative availability of both alcohol and patterns of drug use in the communities being studied.⁴ An earlier study of a subset of the current sample, recruited over a period of 8 years, reported that 66.1% met the accepted criteria for a current substance use disorder.⁵

Patterns of substance use have changed considerably since the most recent comparable study in Sydney, conducted in 1999, which reported rates of alcohol use disorder of 49% in males and 15% in females, and rates of any drug use disorder of 34% and 44%, respectively.² Surveys reported by the Australian Institute of Health and Welfare (AIHW) note the changing patterns of substance use in Australia in the last 20 years, with a decline in the use of heroin and an increase in the use of stimulant drugs, especially methamphetamine.⁶ While methamphetamine users might not seek to use the drug every day in the same way as users of heroin and prescription opioids, the regular use of methamphetamine is associated with the exacerbation of all forms of psychiatric disorder and an increased likelihood of admission to psychiatric hospitals.⁷ A high proportion of homeless populations around the world have mental illness, especially persistent psychotic illness, but also acquired brain injury, mood disorder and trauma related conditions, and comorbid substance use is associated with a less favourable course of those disorders.⁸

The aims of this study were to: (1) provide updated rates and patterns of substance use among people attending clinics conducted in Sydney homeless shelters; and (2) describe the demographic and clinical factors associated with substance use.

Method

A retrospective cohort study of 2498 people who attended mental health clinics conducted in homeless hostels in New South Wales (NSW), Australia, using linked clinical and mortality data. Approval from the relevant research ethics committee was obtained to conduct the study.

Data sources and record linkage

The sample comprised of people who attended psychiatric clinics performed at the three main homeless hostels in inner city Sydney, the Matthew Talbot Hostel run by St Vincent De Paul, the Edward Eager Lodge run by the Wesley Mission, and Foster House run by the Salvation Army, from 17 February 2008 to 19 May 2020. Demographic and clinical variables collected included the diagnoses made by the psychiatrists, patterns of substance use, experience of trauma as a child or adult, problem gambling, being homeless for more than a year, sleeping in the open, type of social security benefit, being the subject of a financial management order, criminal conviction during the study period and having been discharged from a psychiatric hospital to homelessness at the time of initial contact. Deaths were identified using the NSW Registry of Births, Deaths and Marriages to 30 June 2021 and where available, the cause of death was coded according to International Classification of Diseases, 10^th Revision Australian-modification (ICD-10-AM).

Probabilistic record linkage was conducted by the Centre for Health Record Linkage (CHeReL). Upper and lower probability cut-offs for a link were 0.75 and 0.25, and record groups with probabilities between the cut-offs were clerically reviewed.

Data analysis was conducted using SPSS version 28. A descriptive analysis compared clients who did not report substance use with those who reported the use of a single substance, mainly alcohol, and those who reported the use of multiple substances. Pearson chi-square tests were used to identify significant difference and a Bonferroni-adjusted p-value <.05 was used. Significant univariate factors were entered into a multivariable logistic regression model to examine predictors of death and odds ratios and 95% confidence intervals reported.

Around 20% of the sample identified as being of Indigenous heritage, but that was not included in the model as separate ethical review had not been obtained. Model fit was assessed using Hosmer-Lemeshow test for logistic regression (p = .143).⁹

Results

A total of 2041 clinic attendees (81.7%) reported a current or previous substance use disorder, with the most common being alcohol (1545, 61.8%), cannabis (1271, 50.9%), stimulant drugs, mainly methamphetamine (873, 34.9%), and both prescription opioids and heroin (604, 24.2%). About a quarter (575, 23%) reported the use of a single substance, and more than half (1466, 59%) reported the use of two or more substances (Table 1).

Table 1.

Type of substance use in whole sample (n = 2498).

	n	%
Any history of substance use disorder	2041	81.7
Type of substance
Alcohol	1545	61.8
Cannabis	1271	50.9
Stimulant	873	34.9
Opioid^a	604	24.2
Benzodiazepine	730	29.2
Number of substances reported
0	457	18.3
1	575	23.0
2	605	24.2
3	430	17.2
4	207	8.3
5	224	9.0

^aIncludes prescription opioids and heroin.

There were significant differences between non-users, those reporting the use of a single substance and polysubstance users in both demographic and clinical factors. Substance users were more likely to be male and to have been employed for more than a year, but were also more likely to receive the Disability Support Pension. Polysubstance users were more likely to have been homeless for more than a year, to report sleeping in the open and to have a criminal conviction in the study period. Substance users were also more likely to report a history of trauma in childhood and adult life, to have a history of acquired brain injury, to report problem gambling, and to have received a clinical diagnosis of personality disorder, either on the basis of an objective pattern of antisocial conduct or because of unstable emotions and a pattern of self-defeating behaviour. There was no significant difference in the patterns of substance use in those with psychotic illness or mood disorder. Not surprisingly, substance users were also more likely to have died within the study period, with polysubstance users more likely to have died from unintentional overdose, and those reporting the user of a single substance, typically alcohol, more likely to have died from cardiovascular disease, malignancy or liver failure (Table 2).

Table 2.

Comparison of demographic and clinical characteristics in non-users, single substance use and polysubstance use.

	Total sample (n = 2498)		No substance use reported (n = 457)		Single substance use* (n = 575)		Polysubstance use (n = 1466)		Sig
	n	%	n	%	n	%	Polysubstance use (n = 1466)		X²	p
Male	2341	93.7	412	90.2^a	543	94.4^b	1386	94.5^b	12.055	.002
Ever married	842	33.7	149	32.6	202	35.1	491	33.5	.800	.670
Children	641	25.7	117	25.6	157	27.3	367	25.0	1.117	.572
Ever employed for more than 1 year	1482	59.3	258	56.5^a	366	63.7^b	858	58.5^a	6.409	.041
Disability support pension	1389	55.6	222	48.6^a	315	54.8^b	852	58.1^b	13.048	.001
Public trustee	199	8.0	43	9.4	43	7.5	113	7.7	1.618	.445
Homeless for more than 1 year	1495	61.8	225	49.2^a	322	56.0^b	948	64.7^c	39.127	<.001
Sleeps in the open	926	37.1	135	29.5^a	191	33.2^a	600	40.9^b	24.117	<.001
Criminal conviction during the study	1646	65.9	185	40.5^a	331	57.6^b	1130	77.1^c	230.695	<.001
History of early or recent trauma	1092	43.7	185	40.5^a	207	36.0^a	700	47.7^b	25.547	<.001
Personality disorder	945	37.8	119	26.0^a	166	28.9^a	660	45.0^b	78.870	<.001
Acquired brain injury	359	14.4	44	9.6^a	93	16.2^b	222	15.1^b	10.583	.005
Problem gambling	290	11.6	37	8.1^a	76	13.2^b	177	12.1^b	7.253	.027
Depression or anxiety	1465	58.6	270	59.1	342	59.5	853	58.2	.328	.849
Psychotic illness	1301	52.1	235	51.4	295	51.3	771	52.6	.372	.830
Discharge from psychiatric hospital to homelessness	540	21.6	94	20.6	104	18.1	332	22.6	5.278	.071
Ten or more contacts with clinic	625	25.0	96	21.0	155	27.0	374	25.5	5.262	.072
Died within study period	324	13.0	42	9.2^a	96	16.7^b	186	12.7^c	12.958	.002
Cause of death	% of deaths
Unintentional overdose	76	23.5	5	1.1^a	13	2.3^a	58	4.0^b	11.222	.004
Suicide							16	1.1	1.858	.395
Injuries					6	1.0	14	1.0	2.633	.268
Cardiovascular					17	3.0^b	12	0.8^a	13.890	<.001
Cancer					15	2.6^b	21	1.4^a	6.697	.035
Digestive					12	2.1^b	9	0.6^a	15.492	<.001
Respiratory	20	6.2	6	1.3	5	0.9	9	0.6	2.188	.335
Other	30	9.3	8	1.8	9	1.6	13	0.9	3.026	.220
Cause unknown or not recorded	63	19.4	13	2.8	16	2.8	34	2.3	.598	.742
Type of substance use
Alcohol			—	—	400	69.6^a	1145	78.1^b	16.368	<.001
Cannabis			—	—	98	17.0^a	1173	80.0^b	697.073	<.001
Stimulant drugs			—	—	41	7.1^a	832	56.8^b	415.477	<.001
Opioid			—	—	17	3.0^a	587	60.0^b	272.605	<.001
Benzodiazepine			—	—	19	3.3^a	711	48.5^b	367.197	<.001

Cells with n < 5 left blank.

Each superscript letter in each row denotes a subset whose column proportions do not differ significantly from each other, using chi-square analysis.

The factors found to be significant in univariate comparisons were entered into a multivariable logistic regression model, which found that any substance use was associated with the likelihood of sleeping in the open, reporting a history of trauma, having a diagnosis of personality disorder and having a criminal conviction during the study period (Table 3).

Table 3.

Multivariable logistic regression of factors associated with substance use.

	Sig	Odds ratio	95% CIs
	Sig	Odds ratio	Lower	Upper
Disability support pension	<.001	1.380	1.157	1.646
Homeless >1 year	.050	1.217	1.000	1.481
Sleeps in the open	.008	1.317	1.076	1.613
History of trauma	.004	1.297	1.087	1.547
Discharge from hospital	.050	1.238	1.000	1.531
Personality disorder	<.001	2.055	1.705	2.476
Criminal conviction	<.001	3.067	2.566	3.666
Death from accidental over dose	.123	1.557	.888	2.731
Death from natural causes	.056	.671	.446	1.011

Sig p < .01.

Discussion

This study of a large sample of predominantly male attenders of clinics performed at hostels for the homeless found that 81.7% had a past or current substance use disorder, and that the majority of those (71.8% of substance users) reported the use of two or more substances. This proportion is higher than the range reported in a study comparing mental disorder among homeless people in high income countries¹ and is much higher than the rates reported in a study of a similar population conducted in 1999² and by the proportion of 8.6% of all homeless people estimated by the AIHW.¹⁰ The higher proportion found in this study might be in part due to the different methodology, which included more detailed observations that were possible through regular contact with the clinics, as most of the sample had been seen more than once and nearly a quarter had been seen on more than ten occasions. However, the results also suggest an increase in the proportion of substance use among the homeless in the 20 years since the last study.

The study confirmed the changes in the patterns of substance use over the last 20 years reported by the AIHW surveys, with a marked increase in the use of stimulant drugs and the abuse of benzodiazepine medication, and also the more ubiquitous use of cannabis. The use of opioids was stable, although the experience from the clinics was of a reduction in the regular use of heroin and an increase in the abuse of prescription opioids, especially among polysubstance users. The study also confirms the high propensity to other forms of addiction in homeless populations, as around 85% of the sample were daily tobacco smokers¹¹ around 12% reported problem gambling¹² and a high proportion had obesity and metabolic disorder, despite low income.¹³

Substance use, especially polysubstance use, was associated with more severe disability, despite being more likely to have been employed for more than a year, with significantly higher proportions in receipt of the Disability Support Pension and being more likely to be homeless for more than a year and to sleep in the open. Those outcomes are an understandable consequence of prioritising substance use over rent. Polysubstance users were also far more likely to have a criminal conviction during the study period, and hence attract a diagnosis of antisocial personality disorder, both because of illegal activity to obtain money for drugs, but also because of drug possession, behaviour arising from intoxication, and being more visible to the police as a result of living in the open.¹⁴

Substance use has been described as both a cause and a consequence of homelessness,¹⁵ with substance use known to result in the loss of employment and tenancy and the breakdown in relationships, and because being homeless is demoralising, increasing the attraction of mind altering substances, and being homeless often results in exposure to a substance using milieu.¹⁵ Because of the high prevalence of substance use, and the high exposure to substance use among clinic attenders, nearly every clinic consultation included some degree of motivational interviewing to help attendees address addictions and other impediments to recovery. There is some evidence that a Housing First strategy results in a reduction in the use of alcohol and cannabis,¹⁶ although the effect on all substance use reported in a review of four controlled trials was inconclusive.¹⁷ The results of this study confirms the need for the integration of mental health, addiction treatment, and housing services.

The main strength of this study is the large sample and the detailed history available from many of the subjects, including in many cases copies of records from other health care agencies. However, the study has a number of limitations. The study only includes people who attended the clinics, and does not reflect the entire homeless population, or even all those who access the hostels. There were only a small number of women, despite women making up a significant and increasing proportion of people defined as homeless, and the sample was too small to report on changes in patterns of substance use in homeless women. The study only reported the overall results from the 12 years of the study, and does not report on temporal trends in substance use in that time, for example, the effect of the more recent restriction in the availability of prescription benzodiazepine medication. A further limitation was the absence of housing data including information on patterns of substance use during the Covid period, when many of the sample were temporarily housed, which might have shown the effect of a housing first approach on substance use.

Conclusion

This study reports disturbingly high rates of substance use among people who attend clinics at homeless hostels in Sydney, and the high level of disability associated with severe substance use. The results confirm the need for comprehensive programs that address both substance use and other causes of homelessness. Contact with the clinics provided an opportunity for collaborative intervention with the Housing Department and housing advocacy services to secure housing and the integration of treatment for mental illness and substance use.

Footnotes

Acknowledgements

The authors wish to acknowledge the homeless clinic clients and to thank the NSW Ministry of Health for providing access to the NSW ED, ambulatory mental health contact, hospitalisation and mortality data, and the CHeReL for conducting the record linkage.

Author contributions

Kirra Solteberck initiated the study and prepared the first draft. Lauren Staples and Rebecca Mitchell cleaned the data and performed statistical analyses, and reviewed the draft. Nicholas Burns collected data and reviewed the manuscript. Logan Woodman reviewed the manuscript. Olav Nielssen conceived the study, collected data, obtained ethical consent and prepared the final draft. All authors reviewed a near complete draft of the paper.

Disclosure

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Professor Nielssen has received payment for participation in a Jansen advisory board. The remaining authors declare that there is no conflict of interest.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The study was assisted by an Investigator Initiated Research Grant From Janssen.

Ethical statement

ORCID iD

Olav Nielssen

References

Fazel

Geddes

Kushel

. The health of homeless people in high-income countries: descriptive epidemiology, health consequences, and clinical and policy recommendations. Lancet 2014; 384: 1529–1540.

Teesson

Hodder

Buhrich

. Substance use disorders among homeless people in inner Sydney. Soc Psychiatr Psychiatr Epidemiol 2000; 35: 451–456.

Teesson

Hodder

Buhrich

. Alcohol and other drug use disorders among homeless people in Australia. Subst Use Misuse 2003; 38: 463–474.

Fazel

Khosla

Doll

, et al. The prevalence of mental disorders among the homeless in western countries: systematic review and meta-regression analysis. PLoS Med 2008; 5: e225.

Nielssen

Stone

Jones

, et al. Characteristics of people attending psychiatric clinics in inner Sydney homeless hostels. Med J Aust 2018; 208: 169–173.

Australian Institute of Health and Welfare . Alcohol, tobacco and other drugs in Australia. Canberra, ACT: Australian Institute of Health and Welfare, 2023. https://www.aihw.gov.au/reports/alcohol/alcohol-tobacco-other-drugs-australia/contents/summary

Riddell

Nielssen

Butler

, et al. The relationship between amphetamine use, crime and psychiatric disorder among prisoners in New South Wales. Psychiatr Psychol Law 2006; 13: 160–165.

Mullin

Gupta

Compton

, et al. Does giving up substance use work for patients with psychosis? a systematic meta-analysis. Meta-Analysis. Aust N Z J Psychiatr 2012; 46: 826–839.

Hosmer

Lemeshow

. Applied logistic regression. 2nd ed. Hoboken, NJ: John Wiley & Sons, 2000.

10.

Australian Institute of Health and Welfare . Homelessness and homelessness services. Canberra, ACT: Australian Institute of Health and Welfare, 2023. https://www.aihw.gov.au/reports/australias-welfare/homelessness-and-homelessness-services

11.

Power

Mallat

Bonevski

, et al. An audit of assessment and outcome of intervention at a quit smoking clinic in a homeless hostel. Australas Psychiatr 2015; 23: 528–530.

12.

Machart

Cooper

Jones

, et al. Problem gambling among homeless clinic attenders. Australas Psychiatr 2020; 28: 91–94.

13.

Nielssen

Chudleigh

Chen

, et al. Results of a metabolic health clinic at a hostel for homeless men. Australas Psychiatr 2017; 25: 270–273.

14.

Mitchell

Burns

Glozier

, et al. Homelessness and predictors of criminal reoffending: a retrospective cohort study. Crim Behav Ment Health 2023; 33: 261–275.

15.

Thompson

Wall

Greenstein

, et al. Substance-use disorders and poverty as prospective predictors of first-time homelessness in the United States. Am J Publ Health 2013; 103: S282–S288.

16.

Cherner

Aubry

Sylvestre

, et al. Housing first for adults with problematic substance use. J Dual Diagn 2017; 13: 219–229.

17.

Baxter

Tweed

Katikireddi

, et al. Effects of Housing First approaches on health and well-being of adults who are homeless or at risk of homelessness: systematic review and meta-analysis of randomised controlled trials. J Epidemiol Community Health 2019; 73: 379–387.