How-To Guides

Aftercare was associated with lower subsequent suicide risk but higher repeat hospital-presenting self-harm. In this nationwide cohort of 90,413 index nonfatal self-harm episodes in Taiwan, adjusted analyses found that receiving aftercare within 90 days was associated with a 20% decrease in suicide risk (HR=0.80, 95% CI 0.74–0.86) and a 17% increase in self-harm repetition risk (HR=1.17, 95% CI 1.14–1.21). During follow-up, 17,280 patients (19.1%) repeated self-harm over a mean of 2.8 years, and 3,317 (3.7%) died by suicide over a mean of 3.3 years.

Repeat self-harm and suicide were both common after an index hospital-presenting self-harm episode. Among 90,413 patients, 19.1% repeated self-harm during a mean follow-up of 2.8 years, and 3.7% died by suicide during a mean follow-up of 3.3 years. These findings underscore the high-risk period after self-harm presentation to hospital.

The suicide incidence rate was lower in the aftercare-exposed group than in the nonexposed group. The reported rates were 738.1 per 100,000 person-years for the exposed group versus 1,650.1 per 100,000 person-years for the nonexposed group. In adjusted Cox models, aftercare was associated with lower suicide risk (HR=0.80, 95% CI 0.74–0.86).

The incidence rate of repeat self-harm was higher in the aftercare-exposed group than in the nonexposed group. The reported rates were 10,132.0 per 100,000 person-years in the exposed group versus 4,398.3 per 100,000 person-years in the nonexposed group. After adjustment for covariates, aftercare remained associated with a higher risk of repeat self-harm hospital presentation (HR=1.17, 95% CI 1.14–1.21).

The most marked reduction in suicide risk was seen in patients aged 10–24 years. In that subgroup, aftercare was associated with a 36% lower suicide risk (HR=0.64, 95% CI 0.48–0.86). The decrease in suicide risk was also more marked in patients without a history of psychiatric disorders (HR=0.76, 95% CI 0.67–0.86) than in those with such a history.

The association between aftercare and higher repeat self-harm was more marked in younger patients under 65 years and in patients with a history of psychiatric disorders. For patients with prior psychiatric disorders, aftercare was associated with a 26% higher risk of repeat self-harm (HR=1.26, 95% CI 1.21–1.31). Across younger age groups under 65 years, the hazard ratios for repeat self-harm ranged from 1.13 to 1.34.

The aftercare program provided follow-up contact after self-harm, mainly by telephone and supplementally in person. Aftercare workers performed brief suicide risk assessments, offered psychological support, and made referrals to health or social services when appropriate. According to the national guideline, the first contact should occur within 3 days after the self-harm episode, and the service should continue for 3 months with a minimum of 2 contacts per month.

In practice, aftercare often started later than the guideline target of 3 days. Among patients who received aftercare, the median time to first contact was 12 days (range 0–364 days). Most patients who received aftercare had first contact within 90 days (92.4%), and 59.8% of the full cohort received aftercare within 90 days.

Aftercare exposure was treated as a time-varying variable. For patients whose first aftercare contact occurred within 90 days after self-harm, follow-up time before that first contact was counted as nonexposed and follow-up time after that contact was counted as exposed. Patients whose first contact occurred more than 90 days after self-harm, as well as those who never received aftercare, were classified as nonexposed for the analysis.

No. This was an observational cohort study using a nonexperimental design, so the findings show associations rather than proving causation. The authors noted that the exposed and nonexposed groups may not have been fully comparable, although analyses adjusted for multiple potential confounders including self-harm method, socioeconomic variables, physical comorbidity, and psychiatric history.

Yes. The main findings were similar in three sensitivity analyses: one that counted only certified suicides, one that defined valid aftercare exposure as first contact within 180 days instead of 90 days, and one that adjusted for baseline differences using propensity score deciles. Across these analyses, aftercare remained associated with lower suicide risk and higher repeat self-harm hospital presentation.

The authors suggested that the higher rate of repeat hospital-presenting self-harm may reflect increased help-seeking and service engagement rather than treatment failure. They noted that trust built through aftercare may encourage patients to seek hospital care during crises, and that some patients may shift to less lethal self-harm methods that are more likely to result in emergency department attendance. The authors also suggested that lower suicide risk may relate to greater likelihood of receiving treatment after help-seeking and possibly a shift toward less lethal methods, although they stated that future research is needed to clarify these mechanisms.

• This was an observational study, so causal effects cannot be inferred directly.
• The aftercare intervention bundled several components—risk assessment, psychological support, and referrals—so the effects of individual components could not be separated.
• The analysis included only registered self-harm episodes presenting to hospitals, so findings may not generalize to self-harm in the community that never comes to medical attention.
• The database did not identify the specific hospital unit reporting the self-harm event, although most cases were thought likely to have presented to emergency departments.