Stepped care model in mental health

Full title: Stepped care model in mental health

Authors: Rivero-Santana A, Ramos-García V, Torres-Castaño A, García-Pérez L, Herrera-Ramos E, Hernández-Yumar A, Linertová R, Cazaña-Pérez V, Gómez-Ramos AM, Álvarez-Pérez Y, Favaro F, Santos-Álvarez A, Duarte-Díaz A, Rodríguez-Drincourt de Elízaga J, Capafons-Sosa JI, Ramallo-Fariña Y, Sirumal-Rodríguez E, Corujo-Bolaños G, Alonso-Socas I, Rodríguez-Díaz B, Pérez-Celis A, Acosta-Artiles FJ, Serrano-Pérez PG, Badalon-Higueras L, López-Rodríguez JA, Perestelo-Pérez L

Contact persons: Amado Rivero Santana (amado.riverosantana@sescs.es) / Vanesa Ramos García (vanesa.ramosgarcia@sescs.es) / Lilisbeth Perestelo Pérez (lilisbeth.peresteloperez@sescs.es)

SUMMARY

Introduction

Mental disorders constitute one of the major public health challenges worldwide due to their high prevalence and the impact they have on disability and quality of life. The World Health Organization (WHO) estimates that approximately one in eight people worldwide suffers from some form of mental disorder, with depressive and anxiety disorders being the most common.

The stepped care model (SCM) is an organizational strategy aimed at optimizing available resources, personalizing care according to severity, and improving access to effective interventions across the entire clinical spectrum. It also facilitates a collaborative approach among professionals and between primary and specialized care services, promoting a continuous flow of care adapted to the patient’s progress.

The SCM proposes a hierarchical sequence of interventions, beginning with those of lower intensity (e.g., psychoeducation, guided self-help, digital interventions) and progressing to more complex treatments, such as individual psychotherapy or pharmacotherapy, when an adequate clinical response is not observed.

Objectives

To evaluate the scientific evidence on the effectiveness, safety and cost-effectiveness of MAE in mental health, as well as the ethical, legal, organizational and social aspects relevant to its implementation.

Method

Three systematic reviews (SRs) were conducted to evaluate: 1) effectiveness/safety (including SRs with meta-analyses and randomized controlled trials); 2) cost-effectiveness (SRs of economic evaluations); and 3) ethical, legal, organizational, social, and environmental aspects.

In addition, an economic evaluation (cost-effectiveness model) was performed.

Results

Effectiveness and safety

DEPRESSION

Significant effects on the prevention of clinical disorder were observed at 12 months in individuals aged 65 years and older (or living in nursing homes) (k = 2; RR = 0.47, 95% CI: 0.25, 0.89; I² = 0.05%), but not at 24 months (k = 2; RR = 0.64, 95% CI: 0.27, 1.50; I² = 67.8%).

The effect was significant at 3–6 months on symptom reduction (g = -0.25, 95% CI: -0.41, -0.09; I² = 89.9%), treatment response (k = 10; RR = 1.47, 95% CI: 1.23, 1.75; I² = 70%), remission (k = 14; RR = 1.54, 95% CI: 1.26, 1.89; I² = 84.4%), and the mental component of quality of life (k = 7; g = 0.31, 95% CI: 0.12, 0.49; I² = 83%). For clinical variables, the result was only significant in the subgroup with the highest average baseline symptom severity. At 9–15 months, the results remained significantly favorable to the intervention.

There were no differences in hospitalization rates, and the intervention group showed a significantly higher rate of patients receiving antidepressants in the subgroup with the highest baseline severity (k = 8; RR = 1.41, 95% CI: 1.10, 1.83; I² = 92.4%).

ANXIETY

In adults, there were no significant differences in symptom reduction (k = 5; g = 0.00, 95% CI: -0.40, 0.41; I² = 82.3%), remission (OR = 1.15, 95% CI: 0.74, 1.78; I² = 57.8%), or hospitalizations (k = 1; IRR = 0.7, 95% CI: 0.1, 7.0). One study observed fewer psychiatric emergency department visits (0.26 vs. 0.39 per person). Regarding quality of life (mental health), one study found a significant effect at 12 months, while another found no significant difference. In minors, no significant differences were observed in any variable, although a stepped online CBT program (first half of sessions without therapeutic guidance) may be less effective in reducing symptoms and achieving remission of the disorder than a program with guidance via videoconference after each session (not if the guidance is provided via email).

STRESS

Three studies that applied the same program to different populations with significant stress levels showed improvements in stress reduction (g = -0.69, 95% CI: -0.96, -0.43; I² = 47.6%), anxiety (g = -0.38, -0.92, 0.16; I² = 90.7%), and depression (g = -0.47, -0.82, -0.12; I² = 77.6%), but not in quality of life. Another study obtained significant results for these variables at 6 months, but these improvements were sustained only at 12 months for anxiety.

MIXED SAMPLES WITH ANXIETY/DEPRESSION

There were no significant differences in prevention of the clinical disorder (k = 2; RR = 0.80, 95% CI: 0.46, 1.41; I² = 66.4%), reduction of symptoms of depression (k = 6; g = -0.07, 95% CI: -0.18, 0.03; I² = 42.8%) or minor anxiety (k = 6; g = -0.12, 95% CI: -0.23, 0.00; I² = 51.8%), remission (k = 5; RR = 1.09, 95% CI: 0.91, 1.31), quality of life (k = 5; g = -0.05, 95% CI: -0.18, 0.08; I² = 56.5%), hospitalizations (k = 2), or prescription of Psychotropic drugs (k = 5; RR = 1.03, 95% CI: 0.76, 1.38).

SOMATIC ILLNESS AND DEPRESSION/ANXIETY/STRESS

There were small but significantly favorable effects of the intervention on reducing symptoms of depression (k = 8; g = -0.22, 95% CI: -0.36, -0.07; I² = 50.9%), anxiety (k = 4; g = -0.22, 95% CI: -0.41, -0.03; I² = 27.6%), response (RR = 1.29, 95% CI: 1.08, 1.53; I² = 39.3%), remission (k = 5; RR = 1.36, 95% CI: 1.10, 1.68; I² = 52.2%), and the mental component of quality of life (k = 3; g = 0.31, 95% CI: 0.14, 0.48; I² = 0.31). 11.6%) and increased use of antidepressants (k = 3; RR = 2.15, 95% CI: 1.33, 3.49; I² = 72.9%).

POST-TRAUMATIC STRESS DISORDER

In adults, a small study found a strong effect on the prevention of clinical disorder (RR = 0.36, 95% CI: 0.13, 0.97). The effect on symptom reduction was borderline significant (k = 7; g = -0.25, 95% CI: -0.50, -0.00; I² = 78.6%) and significant for quality of life (g = 0.23, 95% CI: 0.07, 0.39; I² = 0.0%).

In children under 18, there were no significant differences in prevention of clinical disorder (k = 4; RR = 0.99; 95% CI: 0.45, 2.20; I² = 2.8%), symptom reduction (k = 4; g = -0.00, 95% CI: -0.21, 0.22; I² = 3.6%), response (k = 2; RR = 0.93, 95% CI: 0.83, 1.04; I² = 17.4%), or remission (k = 2; RR = 1.22, 95% CI: 0.77, 1.94; I² = 7.3%).

OTHER DISORDERS

No significant differences were observed in OCD (one study with adults and another with children), or bulimia nervosa (2 studies). One study with personality disorders observed a similar mean number of hospitalizations per patient (0.5 vs. 0.6, p = 0.937 for the group x time interaction), but with fewer days of stay with the MAE (4.28 vs. 8.44, p = 0.038), as well as fewer emergency department visits (90.1% of patients who reduced them vs. 67.5%, p < 0.001).

SAFETY

Twenty-six studies reported data on suicidal ideation or behavior, all-cause mortality, or other adverse events. No serious events that could be related to the intervention were observed, and when rates were compared between groups, the results were negative. There were no significant differences or they were favorable to the intervention.

Economic evaluation

 The results show that the average cost of MAE in the treatment of depression is lower than the average cost of usual care (with a difference of approximately €133.08 per patient). QALYs, meanwhile, are higher with MAE compared to usual care (specifically, 0.086 additional QALYs per patient), making MAE a dominant (i.e., less costly and more effective) strategy compared to usual care.

Conclusions

Effectiveness and Safety

GENERAL

  • The available evidence on the MAE is restricted to populations with subclinical symptoms or probable or confirmed mental disorder, and no study was identified that evaluated the application of this model in the management of psychological distress resulting from the social determinants of mental health (precariousness, inequality, gender, etc.).
  • The quality of evidence on the effectiveness of the SCM has been assessed as low or very low for most variables and subpopulations, primarily due to the risk of bias in individual studies, the imprecision of estimates, and, in the case of depression, the high heterogeneity of results.
  • The safety of the model has been evaluated in relatively few studies, with no intervention-related adverse events or significantly worse outcomes observed for the SCM (moderate quality).

DEPRESSION

  • A SCM applied to older adults with depressive symptoms but who do not meet the criteria for a clinical disorder significantly prevents the occurrence of the disorder in the following 12 months (low quality).
  • The effect of the SCM on symptom reduction, assessed continuously, is small and of uncertain clinical significance (very low quality).
  • The SCM may produce a relevant relative and absolute increase in the treatment response rate (low quality) and the remission rate of the disorder (low quality), an effect that could be even greater in populations with greater baseline symptom severity.
  • The effect of the SCM on improving the mental component of health-related quality of life is small and of uncertain clinical significance (low quality).
  • The SCM shows no difference compared to usual care in the rate of psychiatric hospitalizations (low quality).

ANXIETY

Adults

  • The SCM shows no difference compared to usual care in reducing anxiety symptoms (very low quality) or in the remission rate of the disorder (low quality).
  • The results on the effect of the SCM on health-related quality of life are scarce and contradictory (low quality).
  • The SCM showed no difference in the rate of psychiatric hospitalizations (very low quality).

Children/adolescents

  • A stepped online CBT program (first half of sessions without therapeutic guidance) was not inferior to a program with email guidance after each session in symptom reduction and disorder remission, but it was inferior when guidance was provided via videoconference (low quality).
  • A three-phase SCM (psychoeducation, CBT, and individualized treatment) did not differ from usual care in symptom reduction, remission, and quality of life (low quality).

DISTRESS

  • In adults with stress symptoms, a two-phase SCM (guided self-help and problem-solving therapy) compared to usual care may improve levels of stress (low quality), anxiety (very low quality), and depression (very low quality), while no differences were observed in quality of life (low quality).

MIXED SAMPLES (ANXIETY/DEPRESSION)

  • The SCM applied to individuals with symptoms of depression and/or anxiety, but who do not meet the criteria for a clinical disorder, does not significantly prevent the occurrence of a clinical disorder in the following 24 months, compared to usual care (very low quality).
  • The SCM showed no differences in the reduction of depression levels (moderate quality), anxiety levels (moderate quality), disorder remission rates (low quality), psychiatric hospitalization rates (very low quality), or quality of life levels (moderate quality).

SOMATIC ILLNESS AND DEPRESSION/ANXIETY/STRESS

  • The SCM applied to individuals with symptoms of depression and/or anxiety, but who do not meet the criteria for a clinical disorder, does not significantly prevent the occurrence of the disorder in the following 24 months, compared to usual care (low quality).
  • The effect of SCM on the reduction of depression and anxiety symptoms, respectively, assessed continuously, is small and of uncertain clinical significance (low quality).
  • However, SCM produces a relevant increase in the treatment response rate (low quality) and remission of the disorder (low quality).
  • The effect of SCM on quality of life is small and of questionable clinical significance (low quality).

POST-TRAUMATIC STRESS DISORDER

Adults

  • In individuals hospitalized for severe physical injury, MAE may reduce the incidence of PTSD in the following year (very low quality).
  • The effect of MAE on symptom reduction and the mental component of health-related quality of life was significant but of small magnitude (very low quality).

Children/adolescents

  • The SCM showed no differences compared to usual care in PTSD incidence (low quality), symptom reduction (moderate quality), response (low quality), and remission (low quality).

OBSESSIVE-COMPULSIVE DISORDER

Adults

  • An SCM applied to individuals with OCD did not reduce symptoms (very low quality) or achieve remission (very low quality).

Children/adolescents

  • An SCM applied to minors with OCD did not reduce symptoms, but was shown to be non-inferior to usual care (moderate quality).
  • No significant differences were observed in treatment response (low quality) or remission (low quality).

EATING DISORDERS

  • In adults with bulimia nervosa, the SCM did not differ from standard care in symptom reduction, abstinence, or remission (low quality). There may be an interaction with the prognostic risk level, resulting in outcomes that are better than or equal to usual care in high-risk patients, but worse in low-risk patients.

PERSONALITY DISORDERS

  • In individuals with personality disorders recently discharged from a psychiatric inpatient unit or psychiatric emergency department visit, a low-intensity intervention (psychoeducation) before potentially accessing more intensive treatments included in usual care did not reduce subsequent hospitalizations (moderate quality), but it did reduce the length of hospital stay (low quality), and increased the rate of patients who reduced their visits to psychiatric emergency departments (low quality).

Economic Evaluation

  • The economic evaluation conducted for this report showed that the SCM, based on a multi-component program, can be a dominant strategy compared to usual care, that is, more effective and less costly.

Ethical, legal, organizational and social aspects

  • The SCM in mental health is widely accepted by healthcare professionals, patients, and caregivers when implemented as a comprehensive, flexible, and multidisciplinary system capable of adjusting the intensity of care to the actual needs of individuals and promoting person-centered care.
  • The feasibility of the SCM depends primarily on structural and organizational conditions, such as stable funding, staff availability and continuity, reduced administrative burden, clear triage and escalation criteria, and access to training and supervision, rather than on the clinical design of the model itself.
  • Without explicit strategies aimed at improving the system’s visibility, the comprehensibility of the model’s operation, and the reduction of structural and stigma barriers, the SCM risks reproducing inequities in access, primarily benefiting those with greater resources, support networks, or health literacy.

Research Needs

No studies have been identified that specifically address the research needs of the SCM, although some studies do address them partially, particularly regarding the integration of digital interventions, coordination among health services, and support for young people on waiting lists. Standardized outcome measures for evaluating the MAE as a service delivery model are lacking. Reference sets have been found for depression, anxiety, and psychological therapies that can guide future evaluations, but specific metrics need to be adapted for this model.

 

DOCUMENTS:

  1. Full report:

 
SESCS report – Stepped Care Mental Health – Nipo.pdf