Abstract
Background:
Alcohol use disorder (AUD) is common and associated with increased risks of suicide.
Aims:
To examine health care utilization prior to suicide in persons with AUD in a large population-based cohort, which may reveal opportunities for prevention.
Method:
A national cohort study was conducted of 6,947,191 adults in Sweden in 2002, including 256,647 (3.7%) with AUD, with follow-up for suicide through 2015. A nested case-control design examined health care utilization among AUD cases who died from suicide and 10:1 age- and sex-matched controls.
Results:
In 86.7 million person-years of follow-up, 15,662 (0.2%) persons died from suicide, including 2,601 (1.0%) with AUD. Unadjusted and adjusted relative risks for suicide associated with AUD were 8.15 (95% CI, 7.86–8.46) and 2.22 (2.11–2.34). 39.7% and 75.6% of AUD cases who died from suicide had a health care encounter <2 weeks or <3 months before the index date, respectively, compared with 6.3% and 25.4% of controls (adjusted prevalence ratio and difference, <2 weeks: 3.86 [95% CI, 3.50–4.25] and 26.4 percentage points [24.2–28.6]; <3 months: 2.03 [1.94–2.12] and 34.9 [32.6–37.1]). AUD accounted for more health care encounters within 2 weeks before suicide among men than women (P=0.01). 48.1% of last encounters were in primary care and 28.9% were in specialty outpatient clinics, mostly for non-psychiatric diagnoses.
Conclusions:
Suicide among persons with AUD is often shortly preceded by health care encounters in primary care or specialty outpatient clinics. Encounters in these settings are important opportunities to identify active suicidality and intervene accordingly in patients with AUD.
INTRODUCTION
Alcohol use disorder (AUD) affects an estimated 5% of adults worldwide and is a leading cause of morbidity and mortality (1–3). AUD has been identified as among the strongest risk factors for suicidal behavior (4–9). Men or women with AUD have been reported to have more than 4-fold risks of completed suicide compared with the general population, after adjusting for sociodemographic differences and other psychiatric and somatic disorders (9). Because of these high risks and its high overall prevalence, AUD has been estimated to account for 20% of all disability-adjusted life years (DALYs) lost due to suicide (10). Little is known about health care utilization patterns prior to suicide in persons with AUD. Such patterns may reveal key opportunities to prevent suicide in this high-risk patient population.
Several prior studies have explored health care utilization before suicide in general populations or patient samples. Most of these studies have reported that a large majority (80–90%) of individuals who died from suicide had a health care encounter in the previous 1 year (11–15), and approximately 30–40% had a primary care encounter in the previous 1 month (11, 16). Health care utilization also varies across different psychiatric disorders and patient subgroups (13, 14). A US study of male military veterans with substance use disorders (either AUD or drug use disorders) reported that 94.6% and 55.6% had a health care encounter within 1 year or 1 month before suicide, respectively (17). However, to our knowledge, no large population-based studies have examined these patterns specifically in persons with AUD.
We conducted a large cohort and nested case-control study in Sweden to examine health care utilization patterns among adults with AUD who died from suicide. Our goals were to: (1) determine the risk of suicide among persons with AUD; (2) provide the first population-based estimates of health care utilization prior to suicide in this patient population; and (3) assess for sex- and age-specific differences. The results may help inform the development of more effective health care intervention strategies to prevent suicide in persons with AUD.
METHOD
Study Design and Population
This study consisted of both cohort and nested case-control designs. First, a national cohort study was conducted of all 6,947,191 persons aged ≥18 years who had lived in Sweden for at least 2 years as of January 1, 2002, as identified in the Swedish Total Population Register. This register contains demographic information for nearly 100% of persons living in Sweden since 1968 (18). Within this cohort, a nested case-control study was conducted of all 2,601 persons who died from suicide during 2002–2015 and who had a registration of AUD within the previous 2 years, as identified using national population registries (as described below). Each of these cases was matched to 10 controls randomly sampled from the general population who had the same birth year, birth month, and sex, and who were still living in Sweden on the respective case’s death date (i.e., index date).
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008. All procedures involving human subjects were approved by the ethics committee of Lund University in Sweden (No. 2013/736). Participant consent was not required as this study used only anonymized registry-based secondary data.
Alcohol Use Disorder (AUD) Ascertainment
AUD was identified using nationwide diagnoses and alcohol-related convictions reported during 2000–2015. First, International Classification of Diseases (ICD) codes for AUD were identified from all primary or secondary diagnoses in the Swedish Hospital, Outpatient, and Primary Care Registries (ICD-8: 291, 303, 357F, 425F, 535D, 571A-D, 980, V79B; ICD-9-CM: same codes as ICD-8, plus 305A; ICD-10-CM: F10 except F10.0, Z50.2, Z71.4, E24.4, G31.2, G62.1, G72.1, I42.6, K29.2, K70.0-K70.9, K85.2, K86.0, O35.4). The Swedish Hospital Registry started in 1964 and initially included all hospital discharge diagnoses from six populous counties in southern Sweden, but was expanded to cover >99% of the national population starting in 1987 and onward (19). The Outpatient Registry started in 2001 and contains outpatient diagnoses from all specialty clinics nationwide. The Primary Care Registry initially included all primary care diagnoses from two populous counties covering 20% of the national population starting in 1998, but was expanded to cover approximately 45% of the national population in 2001 and 75% by 2008 and onward (20). In addition, AUD was identified in all individuals having at least two convictions for drunk driving (law 1951:649) or drunk in charge of a maritime vessel (law 1994:1009), using nationwide data from the Suspicion and Crime Registers.
Suicide Ascertainment
The study cohort was followed up for suicide deaths from January 1, 2002, through December 31, 2015, using nationwide data from the Swedish Cause of Death Registry. This registry includes deaths and ICD codes for cause of death among all persons registered in Sweden since 1960, with compulsory reporting nationwide. All intentional deaths were identified using ICD-10-CM codes X60-X84, and deaths of undetermined intent were identified using ICD-10-CM codes Y10-Y34. Prior studies have suggested that substantial numbers of suicides may be misclassified as deaths of undetermined intent (21, 22), and that such misclassification may be more common in persons with AUD (23, 24). In the present study, intentional deaths and deaths of undetermined intent were analyzed together as the primary outcome and separately as secondary outcomes.
Ascertainment of Health Care Encounters
In the nested case-control study, all health care encounters in inpatient, specialty outpatient, and primary care outpatient settings during the year prior to the index date, regardless of diagnosis, were identified using the Swedish Hospital, Outpatient, and Primary Care Registries (as described above). To exclude encounters that were the direct result of a fatal suicide attempt, hospitalizations were included only if the discharge date preceded the recorded date of death. In addition, all health care encounters were excluded if they occurred on the death date and contained a diagnosis directly related to mortality (i.e., ICD-10-CM R96-R99; “ill-defined and unspecified causes of mortality”).
Other Study Variables
Sociodemographic characteristics and psychiatric disorders that may be associated with AUD, suicide, and health care utilization were examined as adjustment variables. Sociodemographic factors were identified from the Total Population Register and national census data and were 100% complete. These variables included age (continuous and categorical [18–24, 25–34, 35–44, 45–54, 55–64, 65–74, ≥75 years] to allow for a non-linear effect), sex, marital status (married/unmarried), and education level (≤9, 10–12, >12 years) at baseline. Psychiatric disorders included drug use disorders (ICD-8: 304; ICD-9-CM: 292, 304–305, except 305.0 and 305.1; ICD-10-CM: F11-F16, F18–19; Suspicion Register codes 3070, 5010–5012; Crime Register codes for laws covering narcotics [law 1968:64] and drug-related driving offences [law 1951:649]), affective disorders (ICD-8: 296.0–296.8, 300.4; ICD-9-CM: 296A-296E, 296W, 300E, 311; ICD-10-CM: F30-F39, except 32.3), anxiety/phobia disorders (ICD-8: 300.0, 300.2; ICD-9-CM: 300A, 300C; ICD-10-CM: F40–41), psychotic disorders (ICD-8: 291, 295, 297–299; ICD-9-CM: 291–292, 295, 297–299; ICD-10-CM: F20-F25, F28-F29, F32.3, X.5 in F10-F19), personality disorders (ICD-8/9-CM: 301; ICD-10-CM: F60), and other psychiatric disorders (ICD-8: 300.1, 300.5–300.9; ICD-9-CM: 300B, 300F-300H, 300W, 300X, 307B, 307F; ICD-10-CM: F43-F45, F48, F50).
Statistical Analysis
In the national cohort analysis, Poisson regression with robust standard errors was used to determine incidence rate ratios (IRRs) and 95% confidence intervals (CIs) for suicide associated with AUD. Analyses were conducted using three models: (1) unadjusted, (2) adjusted for sociodemographic factors (age, sex, marital status, education level), and (3) further adjusted for psychiatric disorders (as above).
In the nested case-control analysis, Poisson regression with robust standard errors was used to compute prevalence ratios (PRs) and 95% CIs for the prevalence of a health care encounter in specific time intervals before the index date (<2 weeks, 1 month, 3 months, 6 months, 1 year) among cases compared with controls. In addition, generalized linear models with a Poisson distribution, identity link function, and robust standard errors were used to compute prevalence differences (PDs) and 95% CIs for those same prevalences in cases vs. controls. Each of these models was performed both unadjusted and adjusted for covariates (as above). Poisson model goodness-of-fit was assessed using Pearson and deviance tests and was met in each model.
Interactions between AUD and sex or age were examined in relation to health care encounter prevalence on both the additive and multiplicative scale. In a sensitivity analysis, the analyses of health care utilization were repeated after identifying AUD based on any reported lifetime history rather than within the previous 2 years. All statistical tests were 2-sided and used an -level of 0.05. All analyses were conducted using Stata version 15.1.
RESULTS
A total of 256,647 persons (3.7% of the cohort) were identified with AUD during the study period. Table 1 reports participant characteristics in the total cohort, AUD cases, all persons who died from suicide, and all persons who died from suicide and had an AUD registration in the previous 2 years. Compared with the total cohort, persons with AUD or who died from suicide were more likely to be ages 35–64 years, male, or unmarried; have low education level, drug use or other psychiatric disorders; and/or have more frequent hospitalizations or specialty clinic encounters. In addition, AUD cases were more likely to average at least 2 primary care encounters per year.
Table 1.
Characteristics of study participants, 2002–2015, Sweden.
| Total population | AUD | Completed suicide | Completed suicide and AUD | |
|---|---|---|---|---|
| N=6,947,191 | N=256,647 | N=15,662 | N=2,601 | |
| (100.0%) | (3.7%) | (0.2%) | (0.04%) | |
| n (%) | n (%) | n (%) | ||
| Age at baseline (years) | ||||
| 18–24 | 710,230 (10.2) | 21,985 (8.6) | 1,489 (9.5) | 218 (8.4) |
| 25–34 | 1,161,433 (16.7) | 35,040 (13.6) | 2,243 (14.3) | 388 (14.9) |
| 35–44 | 1,239,781 (17.9) | 56,503 (22.0) | 3,217 (20.5) | 708 (27.2) |
| 45–54 | 1,188,344 (17.1) | 64,305 (25.1) | 3,307 (21.1) | 753 (29.0) |
| 55–64 | 1,115,880 (16.1) | 53,530 (20.9) | 2,668 (17.0) | 431 (16.6) |
| 65–74 | 739,552 (10.6) | 19,384 (7.5) | 1,502 (9.6) | 92 (3.5) |
| ≥75 | 791,971 (11.4) | 5,900 (2.3) | 1,236 (7.9) | 1 (0.4) |
| Sex | ||||
| Male | 3,403,918 (49.0) | 189,634 (73.9) | 10,944 (69.9) | 1,974 (75.9) |
| Female | 3,543,273 (51.0) | 67,013 (26.1) | 4,718 (30.1) | 627 (24.1) |
| Education level (years) | ||||
| ≤9 | 1,859,915 (26.8) | 82,381 (32.1) | 5,215 (33.3) | 841 (32.3) |
| 10–12 | 2,923,018 (42.1) | 130,035 (50.7) | 7,333 (46.8) | 1,363 (52.4) |
| >12 | 2,164,257 (31.1) | 44,231 (17.2) | 3,114 (19.9) | 397 (15.3) |
| Marital status | ||||
| Married | 2,996,887 (43.1) | 77,108 (30.0) | 5,083 (32.4) | 647 (24.9) |
| Unmarried | 3,950,304 (56.9) | 179,539 (70.0) | 10,579 (67.6) | 1,954 (75.1) |
| Drug use disorders | 177,641 (2.6) | 65,279 (25.4) | 3,607 (23.0) | 1,235 (47.5) |
| Affective disorders | 1,055,768 (15.2) | 101,479 (39.5) | 7,035 (44.9) | 1,458 (56.1) |
| Anxiety/phobia disorders | 760,977 (11.0) | 83,671 (32.6) | 4,335 (27.7) | 1,163 (44.7) |
| Psychotic disorders | 156,581 (2.3) | 31,289 (12.2) | 2,488 (15.9) | 560 (21.5) |
| Personality disorders | 84,888 (1.2) | 20,266 (7.9) | 1,727 (11.0) | 443 (17.0) |
| Other psychiatric disorders | 799,988 (11.5) | 61,040 (23.8) | 3,509 (22.4) | 776 (29.8) |
| Hospitalizations (total) | ||||
| 0 | 2,567,361 (37.0) | 42,174 (16.4) | 4,030 (25.7) | 211 (8.1) |
| 1 | 1,263,252 (18.2) | 35,910 (14.0) | 2,435 (15.6) | 255 (9.8) |
| ≥2 | 3,116,578 (44.9) | 178,563 (69.6) | 9,197 (58.7) | 2,135 (82.1) |
| Specialty clinic visits (mean/yr) | ||||
| <1 | 4,612,453 (66.4) | 122,401 (47.7) | 7,603 (48.5) | 883 (33.9) |
| 1 to <2 | 1,333,481 (19.2) | 65,218 (25.4) | 3,466 (22.1) | 629 (24.2) |
| ≥2 | 1,001,257 (14.4) | 69,028 (26.9) | 4,593 (29.3) | 1,089 (41.9) |
| Primary care visits (mean/yr) | ||||
| <1 | 3,852,024 (55.5) | 102,697 (40.0) | 9,462 (60.4) | 1,229 (47.2) |
| 1 to <2 | 1,195,581 (17.2) | 46,554 (18.1) | 2,051 (13.1) | 376 (14.5) |
| ≥2 | 1,899,586 (27.3) | 107,396 (41.9) | 4,149 (26.5) | 996 (38.3) |
AUD and Risk of Suicide
In 86.7 million person-years of follow-up, 15,662 (0.2%) persons in the entire cohort died from suicide (18.0 per 100,000 person-years), including 2,601 persons identified with AUD in the previous 2 years (1.0% of all AUD cases; 80.0 per 100,000 person-years). Unadjusted and fully adjusted IRRs for suicide associated with AUD were 8.15 (95% CI, 7.86–8.46) and 2.22 (2.11–2.34), respectively (Table 2). 1,623 (62.4%) of these deaths were reported as intentional, and 978 (37.6%) as undetermined intent. The fully adjusted IRRs for intentional death or death of undetermined intent associated with AUD were 1.74 (95% CI, 1.63–1.85) and 4.65 (4.13–5.24), respectively. In addition, AUD was associated with a >3-fold risk of non-fatal intentional injury and >1.6-fold risk of non-fatal injury of undetermined intent (Table 2, Adjusted Model 2).
Table 2.
Associations between AUD and suicidal behaviors, 2002–2015, Sweden.
| n (% of AUD cases) | Unadjusted IRR (95% CI) | Adjusted Model 1a IRR (95% CI) | Adjusted Model 2b IRR (95% CI) | |
|---|---|---|---|---|
| Completed suicide | 2,601 (1.0%) | 8.15 (7.86, 8.46) | 6.20 (5.96, 6.44) | 2.22 (2.11, 2.34) |
| Intentional death | 1,623 (0.6%) | 6.07 (5.80, 6.34) | 4.59 (4.38, 4.82) | 1.74 (1.63, 1.85) |
| Death of undetermined intent | 978 (0.4%) | 19.26 (17.97, 20.64) | 14.97 (13.86, 16.17) | 4.65 (4.13, 5.24) |
| Intentional injury | 2,245 (0.9%) | 18.01 (17.07, 19.00) | 18.49 (17.41, 19.64) | 3.18 (2.94, 3.43) |
| Injury of undetermined intent | 2,595 (1.0%) | 2.64 (2.54, 2.75) | 2.46 (2.36, 2.56) | 1.69 (1.61, 1.77) |
Adjusted for age, sex, marital status, and education.
Additionally adjusted for drug use disorders, affective disorders, anxiety/phobia disorders, psychotic disorders, personality disorders, and other psychiatric disorders.
AUD = alcohol use disorder, IRR = incidence rate ratio
Health Care Utilization Prior to Suicide
In the nested case-control analysis, AUD cases were substantially more likely than controls to have had a health care encounter within each time interval before the index date (Table 3). For example, 39.7%, 75.6%, and 93.0% of AUD cases who died from suicide had a health care encounter within 2 weeks, 3 months, or 1 year before the index date, respectively, compared with 6.3%, 25.4%, and 51.8% of controls (adjusted prevalence ratio and difference, <2 weeks: 3.86 [95% CI, 3.50–4.25] and 26.4 percentage points [24.2–28.6]; <3 months: 2.03 [1.94–2.12] and 34.9 [32.6–37.1]; <1 year: 1.44 [1.40–1.47] and 27.2 [25.5–28.9]). Figure 1 shows the backward cumulative prevalence (and 95% CIs) for health care encounters within a given time interval before the index date among AUD cases compared with controls.
Table 3.
Prevalence of health care encounters among AUD cases and controls within specific time intervals before index date.
| Outcome and time interval before index date | AUD | Controls | Prevalence Ratio |
Prevalence Difference |
||
|---|---|---|---|---|---|---|
| n (%) | n (%) | Unadjusted (95% CI) |
Adjusteda (95% CI) |
Unadjusted (95% CI) |
Adjusteda (95% CI) |
|
| Completed suicide | 2,601 (100.0) | 26,010 (100.0) | ||||
| <2 weeks | 1,032 (39.7) | 1,635 (6.3) | 6.31 (5.90, 6.75) | 3.86 (3.50, 4.25) | 33.4 (31.5, 35.3) | 26.4 (24.2, 28.6) |
| <1 month | 1,408 (54.1) | 3,090 (11.9) | 4.56 (4.34, 4.78) | 2.81 (2.62, 3.01) | 42.3 (40.3, 44.2) | 31.3 (28.9, 33.6) |
| <3 months | 1,965 (75.6) | 6,611 (25.4) | 2.97 (2.88, 3.06) | 2.03 (1.94, 2.12) | 50.1 (48.4, 51.9) | 34.9 (32.6, 37.1) |
| <6 months | 2,219 (85.3) | 9,763 (37.5) | 2.27 (2.22, 2.32) | 1.67 (1.62, 1.73) | 47.8 (46.3, 49.3) | 32.0 (30.0, 34.0) |
| <1 year | 2,420 (93.0) | 13,474 (51.8) | 1.80 (1.77, 1.82) | 1.44 (1.40, 1.47) | 41.2 (40.1, 42.4) | 27.2 (25.5, 28.9) |
| Intentional death | 1,623 (100.0) | 16,230 (100.0) | ||||
| <2 weeks | 635 (39.1) | 997 (6.1) | 6.37 (5.85, 6.94) | 3.95 (3.49, 4.47) | 33.0 (30.6, 35.4) | 26.4 (23.7, 29.2) |
| <1 month | 866 (53.4) | 1,900 (11.7) | 4.56 (4.28, 4.85) | 2.86 (2.62, 3.12) | 41.7 (39.2, 44.1) | 31.4 (28.5, 34.3) |
| <3 months | 1,228 (75.7) | 4,097 (25.2) | 3.00 (2.88, 3.11) | 2.03 (1.92, 2.14) | 50.4 (48.2, 52.6) | 35.1 (32.3, 37.9) |
| <6 months | 1,378 (84.9) | 6,075 (37.4) | 2.27 (2.20, 2.33) | 1.65 (1.58, 1.72) | 47.5 (45.6, 49.4) | 31.4 (28.9, 34.0) |
| <1 year | 1,504 (92.7) | 8,384 (51.7) | 1.79 (1.76, 1.83) | 1.42 (1.38, 1.46) | 41.0 (39.5, 42.5) | 26.4 (24.2, 28.5) |
| Death of undetermined intent | 978 (100.0) | 9,780 (100.0) | ||||
| <2 weeks | 397 (40.6) | 638 (6.5) | 6.22 (5.59, 6.92) | 3.76 (3.20, 4.42) | 34.1 (31.0, 37.2) | 26.6 (22.9, 30.3) |
| <1 month | 542 (55.4) | 1,190 (12.2) | 4.55 (4.22, 4.92) | 2.76 (2.45, 3.10) | 43.3 (40.1, 46.4) | 31.3 (27.3, 35.2) |
| <3 months | 737 (75.4) | 2,514 (25.7) | 2.93 (2.79, 3.08) | 2.02 (1.87, 2.17) | 49.7 (46.8, 52.5) | 34.7 (30.9, 38.4) |
| <6 months | 841 (86.0) | 3,688 (37.7) | 2.28 (2.20, 2.36) | 1.71 (1.62, 1.80) | 48.3 (45.9, 50.7) | 33.2 (29.8, 36.6) |
| <1 year | 916 (93.7) | 5,090 (52.0) | 1.80 (1.76, 1.85) | 1.48 (1.42, 1.63) | 41.6 (39.8, 43.4) | 29.0 (26.2, 31.8) |
Adjusted for age, sex, marital status, education, drug use disorders, affective disorders, anxiety/phobia disorders, psychotic disorders, personality disorders, and other psychiatric disorders. (All P<0.001)
AUD = alcohol use disorder
Figure 1.
Backward cumulative prevalence of a health care encounter within a given time interval before index date among AUD cases vs. controls (dashed lines are 95% CIs).
Separate analyses of intentional deaths and deaths of undetermined intent yielded very similar prevalence ratios and differences compared with analyses of these outcomes combined. For example, the adjusted PR and PD for a health care encounter within 2 weeks before the index date were 3.95 (95% CI, 3.49–4.47) and 26.4 percentage points (23.7–29.2) for intentional deaths, and 3.76 (3.20–4.42) and 26.6 (22.9–30.3) for deaths of undetermined intent, compared with 3.86 (3.50–4.25) and 26.4 (24.2–28.6) for these deaths combined (Table 3).
Supplementary Table 1 reports the setting and primary diagnosis for last health care encounters within 2 weeks or 1 year before suicide among AUD cases. Among 1,032 encounters at <2 weeks, 23.1% were hospitalizations, 28.9% were in specialty clinics, and 48.1% were in primary care. Among all outpatient encounters (either specialty or primary care), the majority (55–61%) were for non-psychiatric and non-injury-related medical diagnoses, and <12% were for AUD. Overall similar patterns were found for last encounters within 1 year before suicide (Supplementary Table 1).
A sensitivity analysis in which AUD was assessed based on any lifetime history (rather than the previous 2 years) yielded slightly lower PRs but the overall conclusions were not substantially changed. For example, the fully adjusted PR for a health care encounter <2 weeks before suicide was 3.58 (95% CI, 3.33–3.84) compared with 3.86 (3.50–4.25) in the primary analysis, and for <1 year before suicide was 1.34 (1.31–1.36) compared with 1.44 (1.40–1.47) in the primary analysis.
Interactions
Interactions were explored between AUD and sex in relation to health care encounters <2 weeks before the index date (Supplementary Table 2). Among controls, men had a slightly lower prevalence of an encounter than women (5.7% vs. 8.0%; adjusted PR, 0.78; 95% CI, 0.71–0.87). However, comparing controls with AUD cases, this prevalence increased from 8.0% to 44.7% among women (adjusted PR, 3.05; 95% CI, 2.64–3.52) and from 5.7% to 38.1% among men (adjusted PR, 4.19; 95% CI, 3.75–4.63). The relatively greater change among men resulted in significantly positive additive and multiplicative interactions (i.e., the combined effect of AUD and male sex on the likelihood of an encounter exceeded the sum or product of their separate effects; P=0.01 and P<0.001, respectively). The positive additive interaction indicates that AUD accounted for more health care encounters within 2 weeks before suicide among men compared with women.
Interactions also were explored between AUD and age (Supplementary Table 3). Among AUD cases with a health care encounter <2 weeks before suicide, 12.0% were aged <35 years, 75.7% were 35–64 years, and 12.3% were ≥65 years (median, 51.4 years). The prevalence of a health care encounter increased with age among controls (from 3.6% to 8.2%), whereas AUD cases had a much higher overall prevalence that did not vary by age (39–40% for each group). A positive multiplicative interaction was noted between AUD and younger ages (<35 years), and a negative multiplicative interaction between AUD and older ages (≥65 years) (P=0.03 and P<0.001, respectively). However, no additive interactions were found, suggesting that AUD accounted for similar numbers of additional encounters among younger or older adults compared with those in mid-adulthood.
DISCUSSION
In this large national cohort, AUD was a strong independent risk factor for completed suicide, which was often shortly preceded by health care encounters, especially in primary care or specialty outpatient settings. 39.7%, 75.6%, and 93.0% of persons with AUD who died from suicide had a health care encounter within the previous 2 weeks, 3 months, or 1 year, respectively, which were significantly higher than among controls. Furthermore, AUD accounted for more encounters within 2 weeks before suicide among men than women, despite fewer encounters among men in the general population. Approximately half of all encounters within 2 weeks before suicide were in primary care clinics, and the majority of those were for non-psychiatric diagnoses.
To our knowledge, this is the first study to examine health care utilization patterns prior to suicide in persons with AUD in a large population-based cohort. A US study of 3,132 male military veterans with any substance use disorders reported that 25.4%, 55.6%, and 75.9% had a health care encounter within 7, 30, or 90 days before suicide, but did not include a comparison group nor assess this separately in those with AUD (17). Other studies have explored health care utilization more broadly in general populations. The largest of those included a Danish study of 11,191 persons who died from suicide and 55,955 controls, which reported that 83% and 32% had a primary care encounter within 1 year or 1 month before the index date, respectively, compared with 76% and 19% among controls (11). A South Korean study of 11,523 persons who died from suicide also reported high prevalences of health care utilization (81% in men, 91% in women) within 1 year before suicide, and increasing frequency during the final 3 months (13). A US study of 5,894 persons who died from suicide reported that 83% received health care in the previous 1 year, and 55% did not have a psychiatric diagnosis (12). A UK chart review study of 247 primary care patients who died from suicide also reported significant heterogeneity in health care utilization across different psychiatric disorders (14). However, these patterns have not previously been examined specifically in patients with AUD.
These patterns are clinically important because of the high prevalence of AUD and its known suicide risks. AUD is among the most common mental disorders, with an estimated worldwide prevalence of 5%, and even higher in high-income (8.4%; 95% CI, 8.0–8.9) or upper-middle income (5.4%; 95% CI, 5.0–6.0) countries (1–3). AUD is also one of the strongest reported risk factors for suicidal behavior (4–9). A meta-analysis of 33 cohort studies reported nearly 10-fold risks of completed suicide in adult men and women with AUD compared with the general population (standardized mortality ratio, 979; 95% CI, 898–1065) (7). A more recent meta-analysis of 31 studies with 420,732 participants reported pooled odds ratios of 3.13 (95% CI, 2.45–3.81) for suicide attempt and 2.59 (1.95–3.23) for completed suicide associated with AUD (4). A Swedish cohort study of 7.1 million adults who overlapped with the present cohort found >4-fold risks of completed suicide associated with AUD in either men or women, even after adjusting for depression, other psychiatric disorders, and somatic disorders (9).
In the present study, persons with AUD had more than a 3-fold higher prevalence of health care encounters <2 weeks before suicide compared with the background prevalence. Furthermore, AUD accounted for more such encounters among men, and similar numbers among younger or older adults compared with those in mid-adulthood. Most of these encounters were for non-psychiatric diagnoses. These findings provide further evidence to support universal screening of patients for suicidality and its major risk factors, including depression and AUD. Prior studies have shown that clinical interventions to prevent suicide are effective but remain underutilized (25, 26). Brief screening for depression, AUD, and suicidality is clinically feasible, effective, and can be administered by medical assistants (27–31). Positive screens should trigger further discussion with the clinician and prompt psychiatric follow-up (32). Health care settings that lack a system to follow up positive screens should prioritize the development of such a system to provide effective care for mental health and suicidality (33).
A key strength of the present study was the ability to examine health care utilization patterns using nationwide inpatient and outpatient (including both specialty and primary care) data. This design minimizes potential selection or ascertainment biases, enabling more robust estimates based on a national population. AUD was ascertained using not only nationwide diagnoses but alcohol-related convictions, thus improving ascertainment with data that are independent of the health care system. The results were controlled for sociodemographic factors and other psychiatric disorders, which also were ascertained using highly complete nationwide data.
This study also had several limitations. First, although it included primary care encounters, they were not available with complete nationwide coverage. The Swedish Primary Care Registry had approximately 45% coverage of the national population in 2001, which increased to 75% by 2008 and onward (20). Primary care encounters are therefore under-reported in the present study and their estimated proportion of all encounters should be considered a lower bound. Second, as in other population-based studies, the reporting of suicides involves some misclassification. However, available data on intentional deaths as well as deaths of undetermined intent enabled separate analyses of these outcomes, which showed little difference in health care utilization patterns. Lastly, generalizability to other populations with different socioeconomic contexts and health care systems is uncertain. These findings will need replication when possible in other countries and diverse populations.
In summary, this large cohort and nested case-control study provides the first population-based estimates of health care utilization patterns before suicide in persons with AUD. The results show that AUD is strongly associated with suicide, which is often preceded by health care encounters in primary care and specialty outpatient settings, especially for non-psychiatric diagnoses. Improved uptake of screening and treatment interventions in these settings is a high priority for suicide prevention among persons with AUD.
Supplementary Material
Funding:
This work was supported by the National Institute on Alcohol Abuse and Alcoholism at the National Institutes of Health [R01 AA027522 to A.E. and K.S.]; the Swedish Research Council; and ALF project grant, Region Skåne/Lund University, Sweden. The funding agencies had no role in the design and conduct of the study; in the collection, analysis, and interpretation of the data; or in the preparation, review, or approval of the manuscript.
Footnotes
Conflicts of Interest: None.
Data Availability: The national registry data on which this study was based were analyzed under strict confidentiality agreements with Swedish authorities. Due to ethical and legal concerns, the supporting data (which come from a large portion of the Swedish population) cannot be made openly available. Further information about the data registries is available from the Swedish National Board of Health and Welfare: https://www.socialstyrelsen.se/en/statistics-and-data/registers/.
REFERENCES
- 1.World Health Organization. Global status report on alcohol and health 2018. 2018. https://www.who.int/substance_abuse/publications/global_alcohol_report/en/.
- 2.Rehm J, Shield KD. Global Burden of Disease and the Impact of Mental and Addictive Disorders. Curr Psychiatry Rep 2019; 21(2): 10. [DOI] [PubMed] [Google Scholar]
- 3.Carvalho AF, Heilig M, Perez A, Probst C, Rehm J. Alcohol use disorders. Lancet 2019; 394(10200): 781–92. [DOI] [PubMed] [Google Scholar]
- 4.Darvishi N, Farhadi M, Haghtalab T, Poorolajal J. Alcohol-related risk of suicidal ideation, suicide attempt, and completed suicide: a meta-analysis. PLoS One. 2015; 10(5): e0126870. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Norstrom T, Rossow I. Alcohol Consumption as a Risk Factor for Suicidal Behavior: A Systematic Review of Associations at the Individual and at the Population Level. Arch Suicide Res 2016; 20(4): 489–506. [DOI] [PubMed] [Google Scholar]
- 6.Borges G, Loera CR. Alcohol and drug use in suicidal behaviour. Curr Opin Psychiatry. 2010; 23(3): 195–204. [DOI] [PubMed] [Google Scholar]
- 7.Wilcox HC, Conner KR, Caine ED. Association of alcohol and drug use disorders and completed suicide: an empirical review of cohort studies. Drug Alcohol Depend 2004; 76 Suppl: S11–9. [DOI] [PubMed] [Google Scholar]
- 8.Reutfors J, Brandt L, Ekbom A, Isacsson G, Sparen P, Osby U. Suicide and hospitalization for mental disorders in Sweden: a population-based case-control study. J Psychiatr Res 2010; 44(12): 741–7. [DOI] [PubMed] [Google Scholar]
- 9.Crump C, Sundquist K, Sundquist J, Winkleby MA. Sociodemographic, psychiatric and somatic risk factors for suicide: a Swedish national cohort study. Psychological medicine. 2014; 44(2): 279–89. [DOI] [PubMed] [Google Scholar]
- 10.World Health Organization. Global Status Report on Alcohol and Health 2014. World Health Organization, 2014. [Google Scholar]
- 11.Schou Pedersen H, Fenger-Gron M, Bech BH, Erlangsen A, Vestergaard M. Frequency of health care utilization in the year prior to completed suicide: A Danish nationwide matched comparative study. PLoS One. 2019; 14(3): e0214605. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Ahmedani BK, Simon GE, Stewart C, Beck A, Waitzfelder BE, Rossom R, et al. Health care contacts in the year before suicide death. J Gen Intern Med 2014; 29(6): 870–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Cho J, Kang DR, Moon KT, Suh M, Ha KH, Kim C, et al. Age and gender differences in medical care utilization prior to suicide. J Affect Disord 2013; 146(2): 181–8. [DOI] [PubMed] [Google Scholar]
- 14.Pearson A, Saini P, Da Cruz D, Miles C, While D, Swinson N, et al. Primary care contact prior to suicide in individuals with mental illness. Br J Gen Pract 2009; 59(568): 825–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Andersen UA, Andersen M, Rosholm JU, Gram LF. Contacts to the health care system prior to suicide: a comprehensive analysis using registers for general and psychiatric hospital admissions, contacts to general practitioners and practising specialists and drug prescriptions. Acta psychiatrica Scandinavica 2000; 102(2): 126–34. [DOI] [PubMed] [Google Scholar]
- 16.Hochman E, Shelef L, Mann JJ, Portugese S, Krivoy A, Shoval G, et al. Primary health care utilization prior to suicide: a retrospective case-control study among active-duty military personnel. J Clin Psychiatry. 2014; 75(8): e817–23. [DOI] [PubMed] [Google Scholar]
- 17.Ilgen MA, Conner KR, Roeder KM, Blow FC, Austin K, Valenstein M. Patterns of treatment utilization before suicide among male veterans with substance use disorders. Am J Public Health. 2012; 102 Suppl 1: S88–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Ludvigsson JF, Almqvist C, Bonamy AK, Ljung R, Michaelsson K, Neovius M, et al. Registers of the Swedish total population and their use in medical research. Eur J Epidemiol 2016; 31(2): 125–36. [DOI] [PubMed] [Google Scholar]
- 19.Ludvigsson JF, Andersson E, Ekbom A, Feychting M, Kim JL, Reuterwall C, et al. External review and validation of the Swedish national inpatient register. BMC Public Health. 2011; 11: 450. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Sundquist J, Ohlsson H, Sundquist K, Kendler KS. Common adult psychiatric disorders in Swedish primary care where most mental health patients are treated. BMC Psychiatry. 2017; 17(1): 235. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Ohberg A, Lonnqvist J. Suicides hidden among undetermined deaths. Acta psychiatrica Scandinavica 1998; 98(3): 214–8. [DOI] [PubMed] [Google Scholar]
- 22.Bjorkenstam C, Johansson LA, Nordstrom P, Thiblin I, Fugelstad A, Hallqvist J, et al. Suicide or undetermined intent? A register-based study of signs of misclassification. Popul Health Metr 2014; 12: 11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Lindqvist P, Gustafsson L. Suicide classification--clues and their use. a study of 122 cases of suicide and undetermined manner of death. Forensic Sci Int 2002; 128(3): 136–40. [DOI] [PubMed] [Google Scholar]
- 24.Allebeck P, Allgulander C, Henningsohn L, Jakobsson SW. Causes of death in a cohort of 50,465 young men--validity of recorded suicide as underlying cause of death. Scand J Soc Med 1991; 19(4): 242–7. [DOI] [PubMed] [Google Scholar]
- 25.Hofstra E, van Nieuwenhuizen C, Bakker M, Ozgul D, Elfeddali I, de Jong SJ, et al. Effectiveness of suicide prevention interventions: A systematic review and meta-analysis. Gen Hosp Psychiatry. 2019. [DOI] [PubMed]
- 26.Mann JJ, Apter A, Bertolote J, Beautrais A, Currier D, Haas A, et al. Suicide prevention strategies: a systematic review. JAMA. 2005; 294(16): 2064–74. [DOI] [PubMed] [Google Scholar]
- 27.Posner K, Brown GK, Stanley B, Brent DA, Yershova KV, Oquendo MA, et al. The Columbia-Suicide Severity Rating Scale: initial validity and internal consistency findings from three multisite studies with adolescents and adults. Am J Psychiatry. 2011; 168(12): 1266–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med 2001; 16(9): 606–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Reiss-Brennan B, Briot P, Cannon W, James B. Mental health integration: rethinking practitioner roles in the treatment of depression: the specialist, primary care physicians, and the practice nurse. Ethn Dis 2006; 16(2 Suppl 3): S3-37-43. [PubMed] [Google Scholar]
- 30.Madras BK, Compton WM, Avula D, Stegbauer T, Stein JB, Clark HW. Screening, brief interventions, referral to treatment (SBIRT) for illicit drug and alcohol use at multiple healthcare sites: comparison at intake and 6 months later. Drug Alcohol Depend 2009; 99(1–3): 280–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Babor TF, Del Boca F, Bray JW. Screening, Brief Intervention and Referral to Treatment: implications of SAMHSA’s SBIRT initiative for substance abuse policy and practice. Addiction. 2017; 112 Suppl 2: 110–7. [DOI] [PubMed] [Google Scholar]
- 32.Bauer AM, Chan YF, Huang H, Vannoy S, Unutzer J. Characteristics, management, and depression outcomes of primary care patients who endorse thoughts of death or suicide on the PHQ-9. J Gen Intern Med 2013; 28(3): 363–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Coffey CE. Building a system of perfect depression care in behavioral health. Jt Comm J Qual Patient Saf 2007; 33(4): 193–9. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.