Systematic review of structured care pathways in major depressive disorder and bipolar disorder

Major depressive disorder (MDD) and bipolar disorder (BD) are common mental health conditions worldwide that are associated with significant morbidity and mortality. The burden of these conditions spans across multiple domains, including functional, social, occupational, and overall quality of life [23, 36, 44, 73]. In addition to several years lost to disability, life expectancy is decreased by 7–11 years in patients with MDD and BD [14], which is not only caused by the 20-fold increased risk of suicide [16] in MDD, but also attributable to increased physical illness, such as comorbid metabolic and cardiovascular disorders [9, 17, 47].

While various treatments are available, including different modalities of psychotherapy, pharmacological treatments, and neurostimulation, approximately 1/3 of patients with MDD [22, 54, 67] show limited symptom improvement contributing to high individual suffering, healthcare costs and economic burden [40, 43]. Structured care pathways (SCPs) are an evidence-based treatment algorithm consisting of a series of steps to guide practitioners in the management of patients with a specific condition until symptom remission is achieved. SCPs have been gaining attention in the past two decades in many areas of medicine, including psychiatry [15, 24, 56]. They facilitate the development and implementation of tailored protocols based on ‘up-to-date’ evidence, capacity building, and sustainable change in care delivery to improve quality, safety and services. SCPs provide an optimal infrastructure to implement guidelines and quality standards, decrease unwanted, sub-standard variations in practice, and improve patient satisfaction [13]. Several treatment algorithms for patients with mood disorders with or without comorbidities have been proposed based on existing evidence. However, many of these algorithms have not been examined for their effectiveness or applicability in clinical trials [24, 25, 31, 33, 65].

The objective of this work was to provide an updated systematic review on the effectiveness of SCPs and treatment algorithms for treating individuals with MDD and BD.

MDD and BD are common mental health conditions associated with significant morbidity and mortality [22, 43, 44]. MDD and BD can be challenging to treat andlead to significant healthcare costs and burden to the patients [40, 43], especially if patients experience treatment resistance [22]. SCPs include treatment algorithms consisting of a series of steps and serve as recommendations and/or guidelines for applying evidence-based medicine and continuous monitoring of symptoms through validated outcome measures. This can result in a decrease of sub-standard variations in practice and better detection of non-response or clinical deterioration to improve patient care [5, 13].

In this systematic review, we examined the outcomes of SCPs in patients with MDD and BD with the aim of providing an ‘up-to-date’ overview of SCPs and assessment of their efficacy and effectiveness in individuals with mood disorders. This review provides an updated examination of SCPs in patients with MDD [8, 32, 71] and to our knowledge, is the first review to include SCPs for patients with BD.

Examining characteristics of included studies indicated that only 6 of the included studies examined SCPs for BD, of which one study was a RCT that was limited to patients with bipolar depression [7], and 2 cohort studies with a control group [49, 62], highlighting a clear need for more research to examine the effectiveness of SCPs in BD.

Fifteen of the studies included were observational studies without control groups and 11 were RCTs. While naturalistic studies allow studying treatments in a setting that is close to everyday clinical practice and provide valuable information on feasibility or effectiveness, well-designed RCTs are better positioned to assess treatment efficacy [45]. However, we included cohort studies and observational studies without control groups in this review to provide a broad picture and assessment of SCPs applied in different settings and study designs. It should be noted that the small number of RCTs examining SCPs in mood disorders show the need for more studies, particularly RCTs, to identify findings that are replicated in multiple studies adequately designed for this purpose.

We assessed the quality of the 11 RCTs included in this review using the Cochrane risk of bias tool [30]. Majority of the studies had random sequence generation to minimize selection bias, ITT analysis to minimize the effect of different rates of attrition between groups, and did not have selective reporting. Blinding of participants and care-providers would have been difficult given ethical and patient safety issues that are inherent to a clinical trial. In five studies, outcome assessors were not blinded, which would have been more feasible to include. It should also be noted that allocation concealment was only done in a minority of studies, however, in an open trial with participants and clinicians being aware of the treatment, this may not be as important. It is of note that half of the studies had other potential sources of biases, suggesting that while the RCTs were adequate in their quality, improvements can be made for developing future studies.

The SCPs found in the literature and included in this review were highly heterogeneous, not only with respect to the treatments and sequence of treatments in the algorithms but also in study length. In addition, there was heterogeneity in the scales used to assess symptom improvement, treatment response and remission, as well as heterogeneity in setting (i.e. primary care or tertiary hospitals), and study population for both clinical and demographic variables. Therefore, we performed a qualitative synthesis of the studies included in this review, as it would not be meaningful to quantitatively analyze and synthesize studies with largely varying designs and populations since effects can largely vary depending on setting (i.e. primary care vs. tertiary center) or medical and psychiatric comorbidities [8]. This heterogeneity had also been described in previously published reviews examining treatment algorithms [24, 32, 71]. This variation in treatment algorithms may be related to several aspects, including the fact that they were developed for different patient populations, such as those with comorbid psychiatric and non-psychiatric disorders, and for different care settings. Algorithms also likely differ as they were designed at different points in time with the earliest study published in 1998 [63] and the most recent in 2022 [18]. Evolving evidence and treatment guidelines over time [25, 34] as well as variation in treatment guidelines in different countries potentially contributed to the variation in SCP design. When the algorithms and/or designs were consistent between studies included in this review, this was because they belonged to the same project or used a previously published algorithm. Algorithms consistently applied across studies included those from TMAP [21, 37, 62, 64, 66], DA VINCI [6, 56, 57], IMPACT [20, 35, 70], GAP [7, 51, 52], and PROSPECT [4, 12]. One common pattern we observed was that ECT was offered in one of the last stages in most SCPs, despite previous studies demonstrating its treatment efficacy, cost-effectiveness and safety [19, 53].

With respect to clinical outcomes, more than half of the studies reported SCP treatment being superior to TAU in remission rates [4, 12, 29, 52], treatment response [12, 20, 21, 29, 35, 37, 49, 70], and change of symptom scores [21, 62, 66]. It is of note that no studies reported superiority of TAU over SCP for both MDD and BD, although one study reported superiority of stratified care compared to SCP [18]. SCP treatment was also found to be superior to TAU in decrease of suicidal ideation [4, 12], anxiety [35], alcohol consumption [56], and patient satisfaction [6]. When comparing to baseline (in the absence of a control group), SCP treatment was shown to decrease alcohol consumption [57] and improve community functioning [64] as well. Collectively, these findings indicate that SCPs may be more efficacious in treating depression and BD compared to TAU. However, the available evidence is mixed and inconsistent and more studies are required to clearly and comprehensively ascertain specific benefits of SCPs.

The heterogeneity of studies made it difficult to determine specific characteristics of the algorithms contributing to their effects. This was also mentioned in previous reviews of standardized treatment algorithms [24, 32]. Based on our inclusion criteria, the SCPs reviewed shared offering a structured algorithm that if properly adhered to ensured that patients received adequate trials of pharmacotherapy and/or psychotherapy and were closely monitored, and treatment was escalated when needed. These factors may have contributed to the favorable patient outcomes in SCPs. Indeed, previous studies have noted that the benefit of SCPs may primarily result from the structured protocols and mandatory assessment of treatment response rather than specific details of the algorithms [8, 32]. In this regard, several studies have examined the effect of measurement-based care, which focuses on using quantitative methods to monitor symptomatic improvement [29, 38, 59]. Also, a recent review noted that change in pharmacological agents did not affect the rate of remission after 2 antidepressant trials [8] in STAR*D, potentially suggesting that the structure of the SCPs including measurement-based care may be a main contributor to their effect, especially in patients with treatment resistant depression, who do not reach remission with 2 or more consecutive trials of antidepressants [55].

Four studies examined the feasibility of SCPs, where their algorithms were found to be applicable to the majority of the screened patients [3, 10, 46]. One study found that nearly all of those who received the SCP treatment were able to adhere to the treatment [41]. However, the majority of studies reported higher dropout rates in the SCP group compared to TAU [4, 7, 20, 35, 51, 52] and two studies showing lower dropout in the SCP group compared to TAU [6, 56]. Both studies with lower dropout in the SCP used the DA VINCI algorithm for treatment of patients with comorbid AUD and depression. Multiple factors of the DA VINCI algorithm may explain this success in decreasing dropout such as its multidisciplinary approach and the combined treatment of AUD and depression which may be more successful in retaining patients than treating either of these disorders. Overall, there was a large variation in dropout rates in both SCP and TAU groups. We were not able to identify specific and/or consistent factors related to study design or algorithm that were associated with dropout rates. Several studies have reported that medication side effects occurred as adverse events during the study [1, 3, 7, 10, 11, 46, 52, 60, 62, 63, 72]. SCPs often included medications with less tolerable side effects, including the thyroid hormone, MAOIs and mood stabilizers as monotherapy or as adjuncts in later algorithm steps, which may have contributed to the higher dropout rates. Furthermore, several studies had included medication dose increases as a step in the algorithm. This approach may have resulted in higher daily medication doses in SCPs, which in turn may have contributed to higher dropout rates secondary to adverse effects. With only one RCT not performing ITT analysis, it is difficult to ascertain how using ITT to account for dropout may affect overall results. It is of note that we did not find a consistent pattern of dropout rates and treatment response or symptomatic improvement. With respect to cost, it was found that SCP have higher cost effectiveness [42] and lower cost per remission [52] than TAU, supporting previous evidence suggesting that standardization of treatment steps and monitoring of treatment response can result in decreased cost of treatment [8].

This review has limitations that should be considered in the interpretation of its findings. It is important to note that SCPs can be defined differently from how they were defined in our study. While certain stepped care, collaborative care, and treatment algorithms fit our definition of SCPs, previous review papers using similar search terms have applied different selection criteria and thereby included different studies [24, 32, 71]. We further focused our review to studies designed with the aim of quantitatively evaluating the performance of a SCP, excluding studies that were designed for a different purpose, such as identifying biomarkers of treatment response. Also, we were mindful of the large heterogeneity in study design and interventions and decided to provide an updated summary rather than a quantitative synthesis (i.e., a meta-analysis). A meta-analysis examining the effectiveness of SCPs will be important in the future, especially to potentially inform clinical or policy recommendations, as more RCTs examining SCPs become available. In addition, our review included observational studies without control groups, which might limit the level of evidence presented. However, by evaluating a broad range of studies, including cohort studies, observational studies and RCTs, we were able to provide a broader assessment and overview of the application of SCPs in different settings. Also, there was a limited number of previous studies examining SCPs for patients with BD, however, accumulating evidence suggests potential effectiveness of SCPs in this population. More studies examining SCPs in patients with BD and MDD, especially with appropriate control groups, would be beneficial to further elucidate the effectiveness of SCPs and/or specific components of SCPs. Finally, we limited our search to 3 databases and only included peer-reviewed articles published in English, which limits the scope of this review. Future reviews performing a broader search of more databases may provide deeper insight into this topic.

The findings of this systematic review suggest that SCPs are equally or more effective as TAU in the treatment of mood disorders. Evidence indicates that SCPs are potentially superior in certain settings, however, further studies are required to establish and confirm this, particularly for patients with BD, before specific recommendations can be made. Future studies should also specifically examine factors contributing to dropout and effectiveness to inform the development and implementation of more effective SCPs for patients suffering from mood disorders. In addition, identification of pragmatic clinical and biological markers to guide the use of SCPs may improve success and may inform integration of individualized medicine approaches and SCPs.