Mental health problems are most often identified and managed in primary care,1 with anxiety and depression accounting for the majority of presentations.2 The nature of these presentations and the illness trajectory varies widely.3 Managing both the volume and heterogeneity of mental health problems that patients present with is a substantial challenge, and both undertreatment and overtreatment are common.4
Stepped-care approaches aim to address this treatment mismatch by delivering the least intensive intervention effective for the individual's level of need. They have been shown to be effective for the treatment of diagnosed depressive and anxiety disorders,5, 6 but less evidence supports the efficacy of stepped care in preventing disorder onset in individuals with subclinical symptoms.6, 7, 8, 9, 10 Furthermore, little is known about whether stepped care reduces the severity of symptoms for people across the severity spectrum.11 More fundamentally, it is unclear how to allocate people to an initial step. There are two approaches to this allocation. In one approach, everyone requiring care is allocated to the lowest intensity treatment and sequentially stepped up to higher intensity treatments if their symptoms do not improve. In the other approach, the initial level of care is matched to the individual's current symptom severity. An alternative approach might be to match care on a needs basis, so that needs are determined using current symptoms, illness trajectory, and other psychosocial factors. This alternative approach, also referred to as matched, targeted, or stratified care, has the potential to considerably reduce time and distress for affected individuals and to enhance engagement in care.5, 12
Research in context
Evidence before this study
Two meta-analyses have compared the efficacy of stepped care with the efficacy of usual care for the treatment of depression, with different results; one showed stepped care to be slightly more efficacious than usual care whereas the other one showed no difference between both groups. One of these meta-analyses also examined the efficacy of stepped care in the treatment of anxiety (found to be more effective than usual care) and prevention of anxiety or depression (found to be no different to usual care). Both meta-analyses highlighted the diversity of stepped-care interventions tested, including the approach to allocating patients to treatment intensity. We searched PubMed and the Web of Science collection of databases for randomised controlled trials (RCTs) published from database inception to March 3, 2020, using the search terms “stepped care” AND (“primary care” OR “general practice” OR “family medicine” OR “family practitioner”) AND (“anxiety” OR “depress*” OR “mental health”). We identified 12 articles that had not been in previous meta-analyses, which reported outcomes of 11 RCTs comparing stepped care with usual care. All but one RCT (which did not provide detailed description of the intervention or comparator group) were of relatively high quality, notwithstanding the potential bias associated with non-blinding of participants that is inherent in trials of non-pharmacological interventions. Five RCTs required participants to meet diagnostic criteria or report symptoms indicative of probable diagnosis, whereas six RCTs also included possible subthreshold symptoms. The predominant approach to treatment allocation was to require all participants to commence at the lowest intensity treatment (six RCTs). Stratification criteria were described in only two RCTs, with treatment allocation dichotomised according to current symptom severity (ie, participants with a score under a particular cutoff were allocated to step 1 and those with a score above this cutoff commenced treatment at step 2). No RCT took a prognostic approach. Evidence of clinical efficacy was mixed. For primary outcomes, five of the 11 stepped-care interventions were favoured over usual care, showing greater improvement in symptom severity or reduced disorder prevalence. Four interventions were associated with improvement in at least one secondary outcome. Treatment adherence varied, with concerns raised that lack of interest in or response to lower-intensity interventions is a barrier to receipt of higher-intensity care when interventions are completed in a stepwise fashion.
Added value of this study
To our knowledge, this study is the largest RCT of a treatment allocation tool for depression and anxiety in primary care. We applied minimal exclusion criteria and included primary care patients with symptoms across the severity spectrum, thus enhancing generalisability to the broader primary care population. Our primary analysis showed that our model of treatment allocation, comprising symptom feedback, treatment priority setting, and prognosis-matched treatment recommendations, resulted in more rapid improvement in psychological distress than usual care. Among people allocated to more intensive intervention, supplementary analyses suggested a gradient of effect in which symptom reduction increased with each additional intervention element received.
Implications of all the available evidence
Delivering mental health treatment at an intensity that is appropriate to individuals' needs can improve mental health outcomes. Our findings suggest that this effect can be achieved through use of an online, patient-completed tool that quickly and systematically classifies people to appropriate management options. This approach supports the widescale implementation of a stepped-care model in which patients are initially allocated to a step appropriate to their needs, minimising the burden that both patients and health professionals can face in accessing or identifying appropriate care and allowing appropriate treatment to be made available efficiently. Further efforts to encourage treatment uptake and engagement, and understand the barriers, are required.
In addition to the question of how treatment is initially assigned is the issue of treatment gaps in existing services. In Australia, there has been considerable investment in increasing population access to partially or fully funded psychological services on referral from a general practitioner (GP). These services are considered most appropriate for people with moderate symptoms, whereas low intensity service types are considered more appropriate for people with mild symptoms and high intensity service types more appropriate for people with severe symptoms. A stepped-care approach could help to identify these groups and provide appropriate services.
We developed Link-me, a stepped-care approach that is digitally supported and prognostic, which aims to stratify people with anxiety or depressive symptoms in primary care to match treatment to their needs. Link-me comprises two interrelated components. The first component is a brief patient-completed Decision Support Tool (DST) that draws on an individual's responses to 23 items (assessing psychosocial factors—namely, gender, mental health history and current symptoms, general health, living situation, financial security) to predict their anxiety and depressive symptom trajectory over the next 3 months and stratify them into one of three prognostic groups (minimal/mild, moderate, or severe; appendix p 3). The second component is a recommendation for treatment pathways (low-intensity or high-intensity care), which depend on the predicted symptom severity. Link-me is based on our previously developed Target-D approach that focused on depression and used the predicted severity of a patient's depressive symptoms at 3 months to match them to either an online programme (low intensity care) or to nurse-led collaborative care or so-called care navigation (high-intensity care).13, 14, 15 In Link-me, we refined the approach by updating the prognostic model for depression and adding one for anxiety, providing a broader range of low-intensity treatment options, expanding the care navigation workforce to include other health professionals (ie, health professionals who are not nurses or trained mental health specialists), and reducing the financial barriers to care for care navigation recipients.
In this randomised controlled trial, we tested the primary hypothesis that using the DST to stratify patients into prognostic groups (minimal/mild or severe) and to provide matched treatment recommendations (low or high intensity) would result in greater reductions in psychological distress at 6 months than would usual care plus attention control (UC+). Our secondary hypothesis was that treatment differences would be supported within prognostic groups (subgroup analyses). Although the DST stratifies people into one of three groups (minimal/mild, moderate, or severe), the moderate prognostic group was excluded from the trial because services are already available for this group (appendix p 6).