Nishani Health

Three studies behind the loop.

What structured follow-up changed in three published trials, and how large the differences were.

Mental illness costs the world more than heart disease, and nearly twice what cancer does. 1

Every dot is $100 billion of economic output lost between 2011 and 2030. One model for all five disease groups.

$16.3T

$15.6T

$8.3T

$4.8T

$1.7T

Mental illness

Heart disease

Cancer

Chronic lung disease

Diabetes

Cumulative, 2010 US dollars. The Lancet Commission rounds the mental illness figure to $16 trillion.

$4.7T

a year: the economic value of health lost to mental disorders worldwide, 20192

“In its acute state, schizophrenia is considered the most impairing of all health conditions.”

World Health Organization, 20253

Structured family support and relapse

73%

decrease in relapse at 12 months

Across 90 randomized trials, 37% with usual care relapsed within 12 months, compared with 10% with structured family support.4

A weekly check-in and the alert it raises

673

less inpatient days over 12 months

With a clinician acting on the alert, the check-in group recorded 37 inpatient days over 12 months, against 710 in usual care.5

Brief therapy and continued contact after an attempt

69%

fewer patients made another suicide attempt within 24 months

Among 120 patients seen after a suicide attempt, 26.7% with usual care made another attempt within 24 months, compared with 8.3% who also received three therapy sessions and personalized letters over two years.6

Sources

Published research about the population, not Nishani results. Caveats that cut against each study’s headline are included.

  1. 1Patel V et al. The Lancet Commission on global mental health and sustainable development. Lancet 2018;392:1553–98; estimate from Bloom DE et al., The Global Economic Burden of Non-communicable Diseases, World Economic Forum & Harvard School of Public Health, 2011, Table 14 (2010 US$; mental illness projected from its share of DALYs).
  2. 2Arias D, Saxena S, Verguet S. Quantifying the global burden of mental disorders and their economic value. eClinicalMedicine 2022;54:101675. 2019 estimate.
  3. 3World Health Organization. World mental health today: latest data. Geneva: WHO, 2025, p. 10.
  4. 4Rodolico D, et al. Lancet Psychiatry. 2022;9(3):201–210. Rodolico et al. pooled 90 randomized controlled trials covering 10,340 participants, comparing family psychoeducation with treatment as usual rather than with another active intervention. Relapse at 12 months was 10% (95% CI 7–14%) with family psychoeducation against 37% with treatment as usual, an odds ratio of 0.18 (95% CI 0.12–0.27). Confidence in the estimate varies from moderate to low depending on which comparison in the network is read.
  5. 5Komatsu R, et al. Schizophr Res. 2013. Two randomized controlled trials. Komatsu (N=45, 22 vs 23, 12 months) tested a weekly check-in from the patient and a family member feeding an algorithm that raised an alert, and the studied protocol had a clinician act on that alert within 24 hours; the trial is small and its confidence interval wide. Španiel (N=146) found the opposite where nobody acted: the intention-to-treat result was null, clinicians took no action after 61% of alerts, and no intervention after an alert carried HR 10.8 (95% CI 1.4–80.0) for hospitalization. Acting on the alerts is what separated the outcomes.
  6. 6Gysin-Maillart A, et al. PLoS Med. 2016;13(3):e1001968. Gysin-Maillart et al. randomized 120 patients admitted to the emergency department of Bern University General Hospital after a suicide attempt to treatment as usual (n=60) or treatment as usual plus the Attempted Suicide Short Intervention Program (n=60): three therapy sessions followed by personalized letters over 24 months. Over 24 months, 8.3% (5 of 60) of intervention participants and 26.7% (16 of 60) of controls made at least one repeat attempt, with 5 attempts in total against 41. The hazard ratio was 0.17 (95% CI 0.07–0.46; p<0.001), which the authors describe as an approximately 80% reduced risk; the 69% shown here is the difference between the two shares, not the hazard ratio. Intervention participants also spent 72% fewer days in hospital (median 29 against 105 among those admitted), a secondary outcome. Caveats: a single site and 120 participants; patients with a psychotic disorder were excluded; dropout at 24 months was 22% in the control group against 7%; the trial was registered after it was completed; and two authors receive royalties from the ASSIP manual.