Racial Differences in Suicidal Behaviors & Post-Attempt Treatment Among Incarcerated Men – Study Notes

Citation & Context

• Lewis, C., Fedock, G., Garthe, R., & Lee, C. (2024). “Racial Differences in Suicidal Behaviors and Post-Suicide Attempt Treatment: A Latent Class Analysis of Incarcerated Men’s Experiences.” Journal of Racial and Ethnic Health Disparities, 11, 3757–3767.
• Administrative data taken from three Midwestern state prisons, covering $2006\text{–}2011$.
• Sample: $N=207$ incarcerated men, $n=345$ documented suicide-attempt incident reports.

Public-Health Background

• Suicide = leading cause of death in U.S. prisons; up to $96\%$ of prison suicides involve men.
• National prison suicide rate ≈ $3$ times that of community-dwelling adults.
• Deaths by suicide in state prisons increased from $168$ (2001) to $311$ (2019); percentage of all prison deaths rose from $5.7\%$ to $8.1\%$.
• Incarceration itself amplifies suicide risk beyond community levels.
• Suicide attempts (SAs) elevate risk of later death by suicide: e.g., $51.8\%$ of deaths occurred after multiple prior attempts.

Known Racial Patterns (Community & Prison Research)

• Prison suicide mortality (per $100{,}000$): White men $29$; Black men $9$; Hispanic men $12$; other $8$.
• Black men disproportionately incarcerated (rate $\approx6\times$ that of White men); nearly $50\%$ arrested by age $23$.
• Community studies: Black men attempt suicide at younger ages (peak $18\text{–}34$ vs White peak $50\text{–}64$) and choose more lethal methods (hanging, firearms).
• Black incarcerated men more likely than Whites to be placed in solitary/segregated housing; single-cell placement is a major SA risk factor.
• Access to post-SA mental-health care racially uneven; White inmates & women more likely to receive services, whereas Black/Hispanic men more often face punitive segregation.

Treatment Standards vs. Reality

• International/US guidelines: immediate health care after SA (on-site medical, transfer to hospital, mental-health evaluation).
• Many correctional policies default to punitive responses (segregation, misconduct reports), which can worsen risk.
• One CA study: >60\% of suicide deaths deemed preventable if treatment had superseded punishment.

Theoretical Lenses

• Deprivation Theory: prison environmental stressors (loss of freedom, isolation, overcrowding, trauma) drive suicidality.
• Importation Theory: inmates “import” pre-existing demographic, cultural, mental-health, and behavioral traits that influence in-prison behavior.
• Contemporary view: combine both to capture institutional & personal risk factors.

Study Objectives

1. Identify racially differentiated patterns of suicide attempts among incarcerated men via Latent Class Analysis (LCA).
2. Examine how these patterns relate to staff responses (health-care requests, segregation, restraint, misconduct reports) and test for racial disparities in responses.
3. Assess associations with age, years already served, and years remaining until earliest release.

Key Variables & Coding

• Demographics: age, race (0 = White, 1 = Black), years served, years to release.
• SA Method (each coded 0/1):
  • Hanging/Suffocation
  • Cutting/Other
  • Lethal Substance Ingestion (razor swallowing, overdose, etc.)
• Location: Segregated housing (1) vs other (0).
• Multiple attempts in 5 yrs: yes (1) / no (0).
• Staff immediate actions (0/1): physical restraint; misconduct report; call to prison medical facility; call to civilian hospital.
• Final disposition (0/1): placement in segregation; transfer/stay at civilian hospital.

Analytic Approach

• Descriptive & χ² tests for race differences in each SA indicator.
• LCA (Mplus 8.1) iteratively fit 1–4 classes; selected model via lowest BIC/aBIC, AIC support, high entropy (≥$0.97$).
• 3-step auxiliary model: race predicting class (ORs with $95\%$ CIs).
• Covariate analysis: age, years served, years left by class (χ² difference tests).
• Distal-outcome comparisons: class → staff responses; also χ² by race.

Results — Descriptive Profile

• Mean age at first/only SA: $30.43\pm8.80\text{ yrs}$ (range $19\text{–}61$).
• Racial composition: $116$ White (56%), $89$ Black (43%).
• Time served to date: $M=5.79\pm5.18$ yrs (range $0\text{–}23$).
• Time remaining: $M=20.06\pm23.49$ yrs (range overdue release → $97$ yrs).

Race-Specific SA Indicators (bivariate)

• Hanging/Suffocation method: Black $49.4\%$ vs White $25.9\%$ (χ² = $12.14$, p<0.001).
• No race difference in: multiple attempts (~$28\%$ each), cutting method, lethal-substance method, segregated-housing location, staff restraint, misconduct report, final segregation, civilian-hospital stay.
• Medical help disparities:
  • Prison medical facility called: White $97.4\%$ vs Black $88.8\%$ (χ² = $6.34$, p<0.05).
  • Civilian hospital assistance requested: White $75.9\%$ vs Black $60.7\%$ (χ² = $5.46$, p<0.05).

Latent Class Solution (3 Classes)

1. Lethal Substance Class (≈$30.9\%$)
   • High: substance ingestion/overdose
   • Low: segregation location, multiple attempts
   • Moderate: categorised as SA by facility
2. Hanging/Suffocation Class (≈$36.7\%$)
   • High: hanging method, segregated housing, official SA label
   • Low: multiple attempts
3. Cutting/Other Class (≈$32.4\%$)
   • High: cutting/head-banging
   • Moderate: segregation, multiple attempts, SA label

Race Predicting Class Membership

• Black men vs White men:
  • Higher odds of Hanging vs Cutting: $OR=3.24\,(95\%\;CI=1.61\text{–}6.53)$.
  • Lower odds of Lethal-Substance vs Hanging: $OR=0.37\,(95\%\;CI=0.19\text{–}0.72)$.

Covariate Differences by Class

• Age: Hanging class youngest ($M=28.67$ yrs) < Lethal ($32.12$) < Cutting ($33.64$); χ² tests significant.
• Years Served: Cutting class longest ($7.51$ yrs) > Lethal ($5.18$) ≈ Hanging ($4.68$).
• Years Left to Release: no significant class differences.

Staff Responses by Class (Distal Outcomes)

• Physical restraint: highest in Hanging ($73.7\%$) ≈ Cutting ($70.1\%$) > Lethal ($33.7\%$).
• Misconduct reports: Cutting ($37.4\%$) > Hanging ($21.0\%$) ≈ Lethal ($21.9\%$).
• Prison medical facility called:
  Cutting $=100\%$ > Hanging ($92.2\%$) > Lethal ($87.4\%$).
• Civilian hospital requested:
  Hanging lowest ($52.6\%$) < Cutting ($70.3\%$) < Lethal ($86.4\%$).
• Final placement in segregation: Cutting ($56.8\%$) > Hanging ($46.0\%$) > Lethal ($32.6\%$).
• Final civilian-hospital stay: Lethal ($45.7\%$) > Cutting ($29.9\%$) > Hanging ($18.4\%$).

Interpretation & Discussion

• Clear heterogeneity in SA patterns; method is the primary class driver.
• Black men disproportionately use hanging/suffocation, are younger, and are commonly in segregation at attempt time.
• Post-SA health care inequities: Black men receive fewer calls for medical help and civilian hospital transfers.
• Hanging class receives more physical restraint yet less medical escalation—potentially compounding risk.
• Findings support both Deprivation (segregation exposure) and Importation (age, race) perspectives.

Practice & Policy Implications

• Prohibit or drastically limit segregated/single-cell housing, especially for high-risk/younger/Black men.
• Universal suicide-risk screening at intake and continuously through first incarceration year; adapt instruments to racialized presentations.
• Mandatory, recurrent staff training (e.g., Suicide Is Forever media module) addressing racial bias and appropriate clinical vs punitive responses.
• Multidisciplinary oversight committees to audit SA incidents for racial disparities and quality of care.
• Invest in culturally responsive, gender-sensitive prevention programs; incorporate Black men’s lived experiences and help-seeking barriers.

Research Directions

• Longitudinal, multi-state datasets capturing full sentence span and mental-health variables.
• Qualitative studies on Black men’s perspectives of suicidality & staff interactions.
• Intersectional analyses (race × gender identity × sexual orientation × mental health) on SA risk and treatment.
• Evaluate effectiveness of specific anti-segregation policies & racially tailored interventions.

Study Limitations

• Restricted to three Midwestern prisons and a 5-year window; generalizability limited.
• Healthcare variables = binary proxies (facility/hospital called); quality/extent of treatment unknown.
• Missing data on mental-illness diagnoses, facility crowding, staff demographics.
• Administrative reports may suffer from under-reporting or misclassification bias.

Key Numerical Take-Aways

• Prison suicide deaths climb from $168$ to $311$ (2001→2019).
• Class distribution: Hanging $\approx36.7\%$; Lethal Substance $30.9\%$; Cutting $32.4\%$.
• Black men hanging method prevalence: $49.4\%$ vs White $25.9\%$.
• Odds Black→Hanging vs Cutting $=3.24$; Medical help disparity (facility called) $\approx9\%$ absolute gap.

Representative References (select)

• Carson (2021) BJS prison suicide statistics.
• Fazel et al. (2017) international prison suicide review.
• Hayes (2010) national jail suicide study.
• Kaba et al. (2015) NYC jail mental-health disparities.
• Konrad et al. (2007) IASP prison-suicide prevention guidelines.