A 2014 analysis of 244,699 NYC jail incarcerations between 2010 and 2013 found that solitary confinement accounted for 53.3 percent of self-harm acts and 45.0 percent of potentially fatal self-harm, despite representing only 7.3 percent of admissions (Kaba et al., American Journal of Public Health 2014, PMID 24521238). The Bellevue outposted therapeutic housing unit, which NYC Health and Hospitals opened in April 2026 with 104 beds for Rikers detainees with serious mental illness, is the hospital-based successor to the model NYC tested inside the jails after that finding. The NYC H+H primary-source press release was not reachable at the time of this writing; we will update the piece when the CMS, NYC H+H, and NYC DOC primary sources are confirmed.
The editorial case for the OTHU rests on what the peer-reviewed record already shows about the population it serves and the programmatic predecessor it extends. The effect size is not the claim; the rationale is.
The CAPS model inside the jail was the programmatic predecessor
From December 2013 through March 2015, the NYC Correctional Health Services piloted the Clinical Alternative to Punitive Segregation (CAPS) across two jail units. A 2016 evaluation in the International Journal of Environmental Research and Public Health compared outcomes for 195 patients in CAPS, 1,433 in Restrictive Housing Units, and a within-subject group of 90 patients who experienced both (Glowa-Kollisch et al., PMID 26848667). Among the 90 who experienced both models, self-harm and injury rates were significantly lower on the CAPS unit than on the RHU. The CAPS model swapped solitary confinement for therapy groups, art therapy, and community meetings, at approximately 1.5 million dollars in additional staff investment per 30-bed unit per year.
The CAPS evaluation is observational and the within-subject sample is small. Writer cannot report a within-OTHU effect size from the CAPS data. What the CAPS evidence supports is the direction of the clinical change when the segregation pressure is removed and therapeutic programming is added. That direction is toward harm reduction, and the OTHU moves the same logic from the jail infrastructure to the hospital infrastructure.
The Rikers population profile justifies a hospital-based approach
A 2015 AJPH paper on the 800 most frequently incarcerated NYC jail patients through 2013 found a clinical profile that supports the hospital-based care model. Those 800 patients averaged 21 incarcerations each, with a median stay of 11 days per admission. Serious mental illness was documented in 19.0 percent of the cohort versus 8.5 percent of a random-sample control; homelessness was 51.5 percent versus 14.7 percent; substance use was 96.9 percent versus 55.6 percent. Most top charges (88.7 percent) were misdemeanors (MacDonald et al., AJPH 2015, PMID 26378829). The frequently incarcerated cohort accounted for 18,713 admissions and 129 million dollars in combined custody and health costs. The paper's authors concluded that "tailored supportive housing is likely to be less costly and improve outcomes" for this population, which is the evidence base the OTHU's funding argument rests on.
What the evidence does not let the piece claim
Writer cannot claim that the OTHU will cut Rikers mortality by a specific percentage. No OTHU-specific outcome data is yet public; the first 90 days of operational outcomes are the kind of data typically released on a quarterly or semi-annual cycle by NYC Correctional Health Services and NYC Health and Hospitals.
Writer also cannot claim that every Rikers health disparity runs in one direction. A 2017 Public Health Reports cross-sectional HCV screening study of 10,790 NYC jail inmates between June 2013 and June 2014 reported 20.6 percent HCV antibody positivity overall, with injection drug use carrying an adjusted odds ratio of 35.0 (95 percent confidence interval 28.5 to 43.0) for HCV positivity in the multivariable model. Non-Hispanic Black inmates had a lower HCV antibody prevalence than Hispanic inmates (adjusted OR 2.1 vs Black reference) and non-Hispanic white inmates (adjusted OR 1.7 vs Black reference) in the same screening cohort (Akiyama et al., PMID 28005477). The Rikers HCV finding is an important counter to an assumed-uniform disparity pattern; some specific disease burdens at Rikers are inverted across race groups, and the piece flags that to preserve editorial credibility.
The Rikers population that the OTHU will serve is majority Black by public reporting
NYC Department of Correction public reporting describes the Rikers Island average daily population as approximately 60 percent Black; the 2024 primary-source confirmation from the NYC DOC annual statistics is still pending on this piece's update list. The peer-reviewed outcomes of the interventions apply to the racial distribution of the incarcerated population as a whole, because none of the evaluated papers (Kaba 2014, Glowa-Kollisch 2016, MacDonald 2015) stratify their primary outcome by race in the abstract. A disproportionately Black incarcerated population means carceral-health interventions at Rikers are Black-health interventions in aggregate, which is the structural framing our HRSA reentry reporting used to introduce this coverage lane.
Homer Venters is the continuity voice across the NYC jail-health record
Dr. Homer Venters, MD, is the former Chief Medical Officer of NYC Correctional Health Services. He is senior or co-author on every peer-reviewed paper cited in this piece: Kaba 2014, Glowa-Kollisch 2016, MacDonald 2015, Brittain 2013, Akiyama 2017, and Martelle 2015 on the NYC jail EHR. He has testified in federal cases on jail conditions as a plaintiff's expert. Among the US clinician-researchers who have published at scale on NYC jail health, he is the continuity voice across 15 years and can speak to how the OTHU relates to the CAPS predecessor he helped evaluate. His current institutional affiliation is pending.
Follow the outcomes through NYC Board of Correction public meetings
The NYC Board of Correction publishes semi-annual reports on jail health and deaths, and its public meeting calendar at nyc.gov/bocunit is where mortality and clinical-outcome data first surface. For readers with a loved one detained at Rikers with serious mental illness, the Fortune Society at fortunesociety.org and the Osborne Association at osborneny.org maintain named legal-navigation resources and clinical-transfer advocacy programs. Both are established 501(c)(3) carceral-reentry organizations rather than general-purpose charities; readers have a specific-action path rather than a generic "learn more" gesture.
Update plan
We will update this piece when NYC Health and Hospitals publishes the Bellevue OTHU press release URL, when NYC DOC confirms 2024 Rikers Island demographics, when the first 90 days of OTHU clinical outcomes are FOIL-retrieved, when the Data Collaborative for Justice November 2025 Rikers mental-health report PDF is pulled, and when the NYC Board of Correction's most recent annual jail-death data is confirmed.
