Asthma-related mortality after release from prison: a retrospective data linkage study

Figures & data

Table 1. Characteristics of the full cohort (all adults released from prisons in Queensland 1994–2007) and the case-control sub-study cohort.

	Female	Male	All
Main cohort data
Number of individuals in study	4,976 (12%)	37,039 (88%)	42,015
Median age at index release (IQR)	30 (24 − 38)	29 (22 − 38)	29 (23 − 38)
Identified as Indigenous	1,212 (24%)	6,813 (18%)	8,025 (19%)
Total person time in community in years	31,150	239,511	270,661
Median person time in community in years (IQR)	6.5 (2.8 − 9.2)	6.8 (2.9 − 9.6)	6.8 (2.9 − 9.5)
Median duration of key incarceration in days (IQR)	48 (13 − 120)	81 (24 − 225)	77 (22 − 206)
Has had a drug conviction	1,459 (29%)	10,744 (29%)	12,203 (29%)
Has had a self-harm warning flag	302 (6%)	2,452 (7%)	2,754 (7%)
Died in community during follow up (1)	237 (5%)	2,092 (6%)	2,329 (6%)
Died due to asthma ^2	2 (0.89%)	18 (0.93%)	20 (0.93%)
Case-control sub study
Number of individuals in study	348 (21%)	1,310 (79%)	1,658
Median age at index release (IQR)	31 (24 − 39)	30 (23 − 40)	30 (23 − 40)
Identified as Indigenous	106 (30%)	436 (33%)	542 (33%)
Total person time in community in years	1,687	6,479	8,166
Median person time in community in years (IQR)	4.7 (2.1 − 7.0)	4.3 (1.8 − 7.8)	4.4 (1.9 − 7.6)
Median duration of key incarceration in days (IQR)	31 (9 − 96)	62 (17 − 174)	55 (14 − 158)
Has had a drug conviction	110 (32%)	343 (26%)	453 (27%)
Has had a self-harm warning flag	2 (0.57%)	23 (1.8%)	25 (1.5%)
Died during follow up	174 (50%)	655 (50%)	829 (50%)
Has been identified as having asthma	95 (27.30%)	173 (13.2%)	268 (16.2%)

¹ There were 46 deaths in custody that were excluded from cohort analysis. ² Asthma as underlying or contributing cause of death, percentage represents the percent of deaths.

IQR = interquartile range.

Table 2. Asthma-related mortality (underlying and contributing) the population of adults released from Queensland prisons 1994–2007 standardized to the Queensland general population (1).

	Person-years of observation	Ex-prisoner Deaths^4	Crude asthma related mortality rate^2 (95%CI)	SMR^3(95%CI)
Indigenous	51,793	5	9.65 (4.02 − 23.2)	5.92 (2.46 − 14.3)
Non-Indigenous	218,868	15	6.85 (4.13 − 11.4)	2.90 (1.74 − 4.82)
All	270,661	20	7.39 (4.77 − 11.5)	3.32 (2.14 − 5.16)

¹ The Queensland general population data is restricted to ages 17 years and older to match the ex-prisoner population. Asthma deaths increase with age, so this age restriction has increased the reported crude mortality rates for the general population. ² Unadjusted crude mortality rates are reported in deaths per 100 000 person years, with the 95% confidence interval. ³ Indirect standardized mortality ratio (SMR) of ex-prisoners compared with the age- and sex-matched Queensland general population, with the 95% confidence interval. ⁴ Number of deaths where asthma was identified as the underlying or a contributing cause.

Table 3. Regression model of all-cause, drug-related, cardiovascular and respiratory disease related mortality in case-control sample of adults released from prisons in Queensland 1994–2007.

	All-cause mortality (N deaths = 829)		Drug-related mortality (N deaths = 159)		Cardiovascular disease mortality (N deaths = 204)		Respiratory disease mortality (N deaths = 117)
Risk parameter^3	Individual^1	Stepwise^2	Individual^1	Stepwise^2	Individual^1	Stepwise^2	Individual^1	Stepwise^2
Age at index release (in decades)	1.41 (1.29-1.55)	1.36 (1.26-1.47)	0.62 (0.48-0.80)	0.55 (0.43-0.72)	2.80 (2.12-3.71)	2.87 (2.17-3.78)	2.84 (1.92-4.22)	3.40 (2.35-4.93)
Asthma identified as a health issue	1.86 (1.40-2.47)	1.60 (1.26-2.03)	2.50 (1.40-4.46)	1.89 (1.11-3.23)	3.20 (1.57-6.51)	3.36 (1.50-7.56)	3.30 (1.63-6.70)	4.64 (2.24-9.60)
Tobacco use (cust/comm)	1.40 (1.10-1.78)		1.44 (0.84-2.44)		0.95 (0.61-1.48)		1.05 (0.58-1.91)
Problematic use of alcohol (cust/comm)	1.56 (1.26-1.94)	1.43 (1.20-1.69)	1.71 (1.08-2.73)		1.49 (0.96-2.31)		2.27 (1.24-4.16)
Drug use ^3	1.55 (1.23-1.94)		3.77 (2.05-6.95)		0.53 (0.36-0.78)		0.74 (0.43-1.27)
Suicide attempt recorded (cust/comm)	1.65 (1.26-2.15)		1.65 (0.98-2.77)		1.55 (0.88-2.72)		2.08 (1.05-4.15)
Non-drug-related interventions (cust) ^4	0.88 (0.69-1.11)	0.69 (0.57-0.84)	1.00 (0.56-1.78)		1.41 (0.88-2.28)		0.95 (0.51-1.78)
General health problems identified: ^5
Cardiovascular disease	1.63 (1.12-2.37)	1.47 (1.10-1.96)	0.63 (0.20-1.91)		6.22 (3.08-12.58)	2.93 (1.40-6.12)	5.67 (1.66-19.34)
Diabetes	2.05 (1.28-3.30)	2.01 (1.35-3.00)	1.00 (0.29-3.45)		2.92 (1.51-5.62)		1.50 (0.53-4.21)
Cancer	1.60 (0.90-2.84)		1.00 (0.25-4.00)		1.40 (0.44-4.41)		4.00 (0.85-18.84)
Blood borne viruses	1.69 (1.28-2.22)	1.38 (1.11-1.72)	3.40 (1.91-6.05)	2.75 (1.66-4.55)	0.89 (0.51-1.54)		0.74 (0.37-1.47)
Nervous system disease	1.99 (1.31-3.02)	1.55 (1.08-2.22)	1.83 (0.68-4.96)		3.40 (1.25-9.22)	4.19 (1.89-9.25)	7.50 (1.72-32.80)
Circulatory disease	1.49 (0.94-2.35)		0.50 (0.15-1.66)		4.38 (2.03-9.43)		4.00 (1.34-11.97)
Respiratory disease	1.58 (1.25-2.00)		1.77 (1.09-2.86)		1.67 (1.03-2.69)		2.77 (1.47-5.22)
Injury and poisoning	1.45 (1.17-1.78)		2.36 (1.48-3.77)		1.49 (0.98-2.26)		1.75 (1.01-3.03)
Mental illness: any diagnosis or psychiatric admission	1.82 (1.45-2.28)	1.53 (1.27-1.84)	3.60 (2.19-5.91)	2.96 (1.86-4.72)	1.85 (1.16-2.96)	2.36 (1.41-3.93)	2.77 (1.47-5.22)	2.37 (1.17-4.78)

Abbreviations: “cust” refers to risk reported during any custodial period; “comm” refers to risk reported while participant was in the general community prior to last release. ¹ Individual models included age as a co-variate, noting that the matched data controls for sex, indigenous status and person-time exposed. ² The stepwise selection model is a single model that includes all risks with a listed value, the algorithmic selection process as detailed by SAS logistic regression documentation (SELECTION = STEPWISE). ^1,2 Results are obtained from a logistic regression with the first number representing the odds ratio point estimate followed by the Wald 95% confidence intervals in parenthesis. ³ Drug use includes use, possession, tested positive to, enrolled for therapy, detoxification or treatment-program for drugs. ⁴ Non-drug-related interventions is a flag that covered any intervention not related to drug-use for which the prisoner was enrolled during the prison stay. ⁵ General health problems that were identified within the prisoner health records for any period prior to end of follow-up.

GBD 2015 Chronic Respiratory Disease Collaborators. Global, regional, and national deaths, prevalence, disability-adjusted life years, and years lived with disability for chronic obstructive pulmonary disease and asthma, 1990-2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet Respir Med. 2017; 5(9):691–706. doi:10.1016/S2213-2600(17)30293-X.

 Web of Science ®Google Scholar