Brief CBT Halves Non-Suicidal Self-Injury During Inpatient Stays — And Distress Tolerance Is the Mechanism

Non-suicidal self-injury occurs in inpatient psychiatric units at rates that clinical teams often treat as near-inevitable — a byproduct of congregating highly distressed patients in high-stress environments. This RCT challenges that assumption directly. A brief, structured cognitive-behavioral protocol reduced NSSI occurrence during the inpatient stay by nearly two-thirds. The mechanism — improved distress tolerance — was not only confirmed but quantified.

The Study Design

Participants were 200 adults with a documented history of suicide attempt, admitted to an inpatient psychiatric unit and randomized to BCBT-Inpatient plus TAU versus TAU alone. BCBT-Inpatient consists of three to four individual sessions using a manual-based protocol derived from Brief Cognitive-Behavioral Therapy for Suicide Prevention, itself one of the more robustly validated outpatient suicide intervention models. The content targets crisis cognitions, reasons for living, safety planning, and distress tolerance skills — all compressed to be deliverable within the typical inpatient stay of days, not weeks.

The primary NSSI outcome was binary: did the patient engage in self-injurious behavior during the stay? Secondary analyses examined distress tolerance using the Distress Tolerance Scale (DTS) at admission and discharge.

What the Numbers Mean

A reduction from 14.3% to 5.6% is a 61% relative risk reduction in NSSI occurrence. The OR of 0.35 means patients in the BCBT-Inpatient arm were roughly one-third as likely to self-injure during hospitalization compared to TAU. This is a clinically meaningful effect on a behavior that carries significant risk of escalation to suicidal behavior and is notoriously difficult to interrupt in high-acuity settings.

The DTS finding completes the picture. Not only did BCBT-Inpatient reduce NSSI — it did so through the mechanism the model predicts: building the capacity to tolerate acute distress without resorting to self-harm. The r=−0.344 correlation between DTS improvement and NSSI absence was statistically significant and held across both treatment groups. Patients who improved more in distress tolerance — regardless of treatment assignment — were less likely to self-injure. This cross-group finding is important: it suggests distress tolerance is a genuine protective mechanism, not just a BCBT-specific artifact.

The Inpatient Adaptation Challenge

Translating evidence-based outpatient protocols to inpatient settings is notoriously difficult. Patients are in acute crisis, often medicated, frequently involuntary, and available for sessions measured in days. BCBT-Inpatient was specifically designed for this context. The four-session structure prioritizes crisis stabilization over the comprehensive cognitive restructuring that characterizes the outpatient model. Sessions are shorter, more directive, and focused on the immediate presenting crisis rather than distal risk factors.

The clinical staff delivering the intervention in this trial received structured training. The question of whether the protocol can be replicated in units without dedicated training infrastructure — which describes most inpatient settings globally — remains open. The intervention is only useful if it can scale beyond research conditions.

Who Was Left Out

The post-hoc NSSI analysis had data on 188 of the 200 enrolled participants. The distress tolerance subsample was smaller still: 53 BCBT-Inpatient and 51 TAU participants completed DTS at both time points. Small subsample sizes inflate the risk of false positive findings in mechanism analyses. The DTS effect requires replication in a prospectively powered sample before the mechanism claim is fully established.

The sample was recruited at a single institution. Inpatient settings vary enormously in staffing models, length of stay, and acuity mix. Effect sizes from single-site trials routinely shrink in multi-site replications.

Clinical Bottom Line

For clinical directors considering protocol additions to inpatient suicide units: BCBT-Inpatient is a 3–4 session structured intervention with RCT evidence for reducing NSSI by roughly two-thirds during hospitalization. The mechanism appears to be distress tolerance enhancement. Implementation requires staff training in the manual-based protocol. The target population — adults admitted after suicide attempt — overlaps extensively with the highest-acuity patients on any acute psychiatric unit. The effect size (OR=0.35) is large enough to warrant consideration even pending multi-site replication.