Pathways to mental health services for young people: a systematic review

1Most psychiatric conditions emerge before the age of 25. Mental illness is the largest contributor to the burden of disability-adjusted life years (DALYs) among young people aged 0–24 in high-income countries and the seventh-highest contributor to DALYs in low- and middle-income countries. Globally, mental illnesses account for a quarter of all years lived with disability (YLDs) in children and youth aged 0–24.

2Despite this heavy burden, many youths with mental health problems remain untreated or face delayed detection, long waitlists and multiple help-seeking contacts before obtaining appropriate care. Such complex ‘pathways to care’ delay treatment. For youths (typically understood as individuals who are within the critical development juncture between childhood and adulthood, i.e., aged between 11 and 25–30 years old), longer durations of untreated illness can have grave impacts on the foundations of their adult lives and can be associated with worse clinical outcomes.

3Pathways to care—defined as the “sequence of contacts with individuals and organizations prompted by the distressed person’s efforts, and those of his or her significant others to seek help, as well as the help that is supplied in response to such efforts”—have been garnering research attention for several years. In the early 1990s, a multinational study by the World Health Organization (WHO) showed that pathways to mental healthcare varied substantially depending on context and resource availability. In regions with access to relatively well-developed mental health services, patients experienced more direct routes from the community to specialized care. However, in areas with few services, patients experienced a wide variety of pathways that often included traditional or faith healers.

4In the field of first-episode psychosis, concern with the adverse consequences of delayed treatment has spurred numerous investigations of pathways to care and barriers to accessing specialized services. In addition to primary care providers and mental health services, help-seeking pathways for psychotic disorders involve diverse contacts like emergency rooms (ERs), social services, the criminal justice system, school counsellors, and religious agencies. Pathways to psychosis services have been known to be influenced by several sociodemographic factors, including gender, age, ethnicity, and socioeconomic status. However, these findings have been inconsistent and their implications for policy and service delivery difficult to assess.

5Sequences of healthcare contacts do not occur randomly, but are influenced by multiple intersecting individual, social, cultural, and systemic factors. Studying pathways to care allows us to identify the loci of barriers and delays to treatment; and key agents in the help-seeking process, including individuals in distress, family/carers, informal contacts (e.g., teachers, employers, web resources, etc.), and formal health services. Such knowledge is crucial for providing timely access to services.

6New youth mental health initiatives, including in but not limited to Australia, Canada, Ireland and the United Kingdom, are striving to make appropriate services accessible early in the course of mental illnesses to mitigate their short- and long-term negative consequences. It has been argued that extant conventional mental health systems are neither youth-friendly nor sufficiently accessible. Young people and their families have described mental health help-seeking as a long, painful, and complicated journey. Though they represent the peak incidence of mental health problems, youths are frequently the least likely to use mental health services and often receive help only when their problems become crises. Their help-seeking efforts may also be impeded by repeated evaluations and difficult transitions, especially between child and adult services.

7Although literature reviews on pathways to care have been conducted in the field of psychosis and across adult mental health disorders, evidence on the different trajectories youths follow to obtain mental healthcare has yet to be synthesized. Such a synthesis is essential if efforts to transform youth mental healthcare are to achieve their ends. Our objective was therefore to conduct a systematic review of literature on young people’s pathways to care for a range of mental health problems.

8The protocol for this systematic review was developed in accordance with the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) and was registered at the PROSPERO Centre for Reviews and Dissemination (ID: 42016039208) in June 2016.

9Search terms were generated by consulting 20 experts in youth mental health across disorders, and a university librarian. We included search terms related to pathways to care; service utilization; help-seeking; mental disorders; and delays to treatment (see online supplementary material for search strategy).

10Relevant studies were identified through searching five electronic databases: MEDLINE (1946 onward), Embase (1947 onward), PsycINFO (1967 onward), HealthSTAR (1966 onward) and CINAHL (1937 onward). Articles were further identified using backward and forward citation mapping of selected articles using Web of Science, and hand searches of journals that had previously published material on pathways (n = 4). The electronic search was conducted in July 2016 and updated in March 2018.

11Two experts independently screened titles, abstracts, and keywords and resolved disagreements by consensus. Articles were included if they were peer-reviewed; were written in English or French; and reported quantitative findings. To be selected, studies had to focus on youths’ individual trajectories to seeking or receiving treatment for mental health or substance use at any establishment, regardless of the presence or absence of a formal diagnosis. The mean age of study participants had to be between 11 and 30 years (so as to include the largest possible range of definitions of ‘youth’ used in pertinent literature). Alternatively, at least 50% of a study’s sample had to be within that age range. We excluded studies of youths with chronic physical health conditions or a primary diagnosis of intellectual disability. Full texts were obtained for all potentially relevant studies. Two reviewers independently screened the full text of each article to check whether it met inclusion criteria.

12The authors of six studies were contacted for additional information to determine their eligibility. Of these, three authors responded and provided data that had not appeared in the original studies, which were then included in our review.

13A data extraction sheet was created and refined following pilot testing on ten randomly selected included studies. Two reviewers independently extracted and compared data from all included studies and resolved disagreements by discussion.

14We extracted data on participant demographics, study design, instruments used, study setting, healthcare context, pathways to care, and measures of treatment delay. If needed, authors were contacted for clarifications or missing information.

15The two reviewers also independently ascertained the quality of each included study using a rating scale adapted from the Newcastle–Ottawa Quality Assessment tool, which had been used in a systematic review on pathways to care in first-episode psychosis (see online supplementary material).

16The electronic search yielded 17,381 publications, including 1454 from the March 2018 search update. Hand searching yielded another 45 articles. After duplicates were removed, 11,524 studies remained. Initial title and abstract screening identified 845 potentially relevant studies for full-text screening. Of these, 45 studies fulfilled the inclusion criteria (see Fig. 1). The main reasons for exclusion were misalignment of studies’ objectives with those of this review, study methodology, language, and participants’ age ranges. Five studies were excluded post hoc because their participants’ age ranges could not be established (n = 3), or for involving the same participants as other included publications (n = 2).

17The characteristics of included studies are summarized in Table 1. The studies were conducted across a wide range of countries (n = 26). Their sample sizes ranged from 15 to 1266 (mean = 203). Twenty-six studies were conducted in services catering to persons with first-episode psychosis. Other study sites were general psychiatric inpatient (n = 3) and outpatient units (n = 9); and specialized services for youths with anorexia (n = 1) and those at risk for psychosis (n = 6).

18We extracted information about the healthcare system in which each study was conducted (Table 2). Many studies described organizational features, including available healthcare tiers (e.g., public/private) and local practices (e.g., preference for traditional healers). Fourteen studies reported allowing open referrals, wherein direct referrals to the services were possible. Two studies described a gatekeeper system where referrals from primary care were required to access mental healthcare. All other studies did not specify their settings’ referral systems.

19Studies differed in the instruments used to ascertain pathways to care. The majority had developed their own interview guide or questionnaire (n = 22) but provided limited to no information on the methodology used to develop the measures or their psychometrics. Semi-structured interview-based instruments included the WHO Encounter Form (n = 14); the Circumstances of Onset and Relapse Schedule for early psychosis (n = 4); the Pathways to Care Schedule (n = 3); and the Basel Screening Instrument for Psychosis (n = 1). One study used the structured Referral Sequence and Problem Interview.

20Irrespective of the instruments used, most studies collected and corroborated information from multiple sources (n = 27). In these cases, individual interviews were supplemented by family/carer interviews and/or chart review. Some studies relied on a single data source—patient interviews (n = 16) or chart information (n = 2). One study used national registry data, which included healthcare contacts and durations of untreated illness.

21Timeframes for delimiting pathways to care, i.e., the start and endpoints of journey into care, differed widely across studies. Startpoints included the onset of symptoms or initial suspected illness (n = 22); 6 months preceding entry (n = 1); lifetime (n = 4); 28 days preceding prodromal symptom onset (n = 1); 48 h prior to admission (n = 1); and first contact with health services (n = 1).

22Endpoints included entry or referral to a specialized service (n = 13); admission to hospital (n = 8); initiation of care (n = 7); and entry to a general psychiatric service (n = 5). For studies that did not specify a timeframe but used the WHO Encounter Form (n = 5), we assumed that instrument’s stated timeframe of 12 months preceding the interview (see Table 2). Other studies did not specify clear start (n = 10) or endpoints (n = 7).

23The focus of this review was on articles that examined individuals’ pathways to care (i.e., sequence or number of help-seeking contacts). Outcome measures included descriptions of full trajectories, or first and last contacts before a specific endpoint. Considered clinically relevant, first and last contacts are often described in pathways to care studies.

24Thirty-five studies described full pathways to care sequences, including the total number and types of contacts in individual participants’ pathways to care. Seven studies described the most common pathway contacts for their sample, in addition to common first and last contacts. Three studies described the most common overall and first contacts along participants’ pathways to care (see Table 3).

25Twenty-eight studies reported the number of contacts before receiving specific services, which ranged from 0 to 15 contacts per participant (with a pooled mean across studies of 2.9 contacts). One study dichotomized pathways into ‘short’ (three or fewer services before referral) and ‘long’ (four or more services). Its authors noted that the number of contacts did not always indicate pathway complexity or length of delay. For example, a pathway with many contacts could reflect appropriate referrals as mental health problems progressed, whereas shorter pathways could reflect repeated contacts with specific services or concurrent use of different services before an appropriate referral.

26Contacts involved in young people’s pathways to mental healthcare were varied and included medical professionals (general practitioners, psychiatrists); non-medical professionals (psychologists, social workers, counsellors, school teachers, rural healthcare workers); informal sources of help (family, friends, employers, colleagues); healthcare institutions (emergency services, inpatient units, walk-in clinics); criminal or justice system (police, prisons, lawyers, courts); traditional or faith-based healers (prayer houses, priests, herbalists, clergy); and technology-enabled contacts (websites, helplines, crisis lines) (Table 3).

27Twenty-nine studies reported the first contact along the pathway to care. In order of frequency, young people’s first help-seeking contacts were general practitioners (14/29); psychiatrists or specialized services (5/29); faith or traditional healers (4/29); ERs/inpatient units (3/29); family or friends (2/29) and social workers (1/29). General practitioners were among the top three most frequent first sources of help in 24 of 29 studies.

28Studies of pathways to care often describe their referral source as the ‘successful contact’, i.e., the contact that resulted in an individual obtaining the service in question. This successful contact is also referred to in some studies as the “last” contact. Twenty-two studies examined referral sources. Of these, eight described the ER/inpatient unit as the most common ‘successful’ referral source. Self-referrals (i.e., referrals made by youths themselves, or by family/carers on their behalf) were the most frequent referral source in six studies. Other prominent referral sources included general practitioners, general hospitals, helplines, and outpatient units.

29Of the 39 studies that measured treatment delay, 23 were from first-episode psychosis settings, and 16 were from other mental health services (see Table 2).

30DUP is defined as the time between the onset of symptoms and the start of appropriate care (operationalized as the commencement of antipsychotic medication or admission to services). Across the 23 studies that reported DUP, mean DUP ranged from 1.5 to 102 weeks and median DUP ranged from 8 to 70 weeks. Of these 23 studies, 10 also assessed ‘help-seeking delays’ (time between the onset of initial symptoms and contact with the first pathway agent) and ‘referral delays’ (time between contact with the first pathway agent and the commencement of treatment at the study setting). Of these, three studies found that help-seeking delays exceeded referral delays; six studies found referral delays to be longer,; and one study found an even split between both delay components. Notably, FEP patients referred to a service for those at risk for psychosis reported longer referral, than help-seeking delays.

31Fifteen studies from a range of mental health settings described the length of treatment delays to their services. Although definitions of DUI varied, most studies conceptualized it as the time between the onset of symptoms and the commencement of treatment at their setting. DUI estimates ranged from 1 week to 45 years (Table 2). Despite our inclusion criteria focusing on young people between the ages of 11 and 30, the upper end of the range for DUI is 45 years. This is because we also included studies in which at least 50% of the included sample was in the age group of interest. Unfortunately, some of these studies did not break down their delay indices by age group (see Table 1 for participant characteristics for each included study.) At the very least, this wide range for DUI is indicative that there are often extremely lengthy delays before the receipt of appropriate treatment. Eight studies divided DUI into help-seeking and referral components. Of these, three studies reported lengthier help-seeking delays and five reported lengthier referral delays.

32Seven studies found that encountering specific pathway agents affected treatment delay. One study found that initial contacts with counsellors or courts led to longer DUPs. Another found that DUP was shorter following referrals from emergency services. DUPs were shorter if the first contact was with general practitioners and when comparing general practitioners to private psychiatrists and psychologists. However, another study reported longer referral delays for persons with FEP following contact with primary care, albeit such contact resulted in fewer negative pathways to care (e.g., emergency or inpatient services). In settings other than psychosis services, contacts with traditional or faith healers or private general practitioners/physicians were notably associated with longer DUIs. Family involvement during help-seeking was associated with shorter help-seeking delays in one study.

33Often, families/friends played a substantial role in the initiation of treatment. In two studies, 70% of participants had sought mental healthcare on the advice of family. One of these studies contrasted this with the much lower rate of individuals deciding on their own to seek services (16%). Families were found to be highly involved at various points along the pathway to care by recommending sources of help, being the most common first source of help, directly initiating contact or being the most common contact. Studies’ methodologies may have influenced their findings. For example, while 12 studies included families/relatives in their definitions of help-seeking contacts, 26 studies only considered professional contacts. Seven studies did not explicitly describe their inclusion criteria for pathway contacts.

34Negative pathways, generally defined as those involving contacts with the criminal justice system, emergency or inpatient units, are associated with poor patient experiences and disengagement; and high costs, despite sometimes resulting in reduced treatment delays.

35A number of studies explored the involvement of police and emergency services along pathways to care. In a study whose entire sample was African–American, over a quarter of participants had at least one contact with police, and police accounted for a fifth of all contacts. In another US study, the pathways of over half the Black participants featured some police involvement, a rate significantly higher than that observed in other ethnicities. In a Canadian study, emergency rooms were four and three times more likely to be the first contact for Asians and other ethnicities, respectively, than for White and Black participants. Overall, emergency services figured prominently as pathway agents across studies and contexts (n = 15).

36Two studies examined the costs associated with various pathways to care. In a Canadian study, pathways to care involving inpatient units were 18.5 times costlier than pathways with no inpatient unit involvement. This was attributable to the greater involvement of police and emergency services with participants who ended up being inpatients. An Indian study demonstrated that the median monetary cost of an individual’s pathway to care was more than half the average family’s monthly income.

37The only three studies that explicitly described being guided by a framework all used Goldberg and Huxley’s conceptual framework. This framework proposes that mental health problems manifest at five levels (from in the community to among those in specialized care), with individuals’ advancement to subsequent levels being checked by selectively permeable filters that pertain to problem recognition (e.g., by general practitioners) and referral (e.g., to specialized care).

38The methodological quality of the studies was mixed (see Table 4 for quality scores). Six studies met over 75% of the quality appraisal criteria; 34 studies met 50–75% of the criteria; and five studies met under 50% of the criteria. Key limitations were insufficient reporting on sample size determination; low participation rates or inadequate differentiation between participants and non-participants; and non-standardized ascertainment of pathways to care.

39Pathways to mental healthcare for youths tend to be complex, with multiple help-seeking contacts, and, sometimes, lengthy delays before appropriate care begins. Across many contexts, general practitioners played a prominent role in the help-seeking process. The role of primary care is notable given the international consensus that integrating mental health services within primary care is essential to address gaps in mental healthcare provision.

40In our reviewed studies, primary care physicians were more frequently among the first help-seeking contacts than a ‘successful’ referral source. To be the first line of mental healthcare, primary care providers must be adequately trained to effectively detect problems, render support, initiate treatment, coordinate with all healthcare tiers, and refer appropriately.

41Across settings, families played an influential role along pathways to care. This highlights the need for including families as pathway agents, something only few studies did. It also indicates that families need to be targeted in outreach efforts to reduce treatment delays for youths. Thus, giving due regard to families is important because familial involvement is known to mitigate the negative effects of and facilitate recovery from many mental illnesses.

42Given the increasing rates of hospitalization and emergency visits among youths with mental health problems, and the high rates of emergency services involvement noted in our review, it is necessary to improve our understanding of the determinants of and trajectories to these endpoints that are associated with high personal and societal costs. Notably, the reviewed studies offer limited insights into what determines which youths follow these negative pathways, barring examinations of ethnicity as a determinant in the case of psychosis.

43Many of the factors leading to fragmented or difficult access to mental health services occur across age ranges. Studies assessing pathways to care in young children and older adults have also reported complex trajectories prior to obtaining services. Notably, however, many mental health systems have attributes that are known to disrupt care specifically for youth; chief among these being the transitions from child–adolescent to adult services. These transitions, often rigid and poorly executed, can lead to disengagement from services and poor clinical outcomes. As such, it may be important for future research to prospectively assess pathways into and through services, and to pay specific attention to how transitions across mental health systems contribute to treatment delays and complicated pathways.

44This review reveals that knowledge on pathways to care in youth mental health is largely driven by first-episode psychosis literature. This is likely due to the field’s focus on reducing the DUP. Despite some disagreements on optimal treatment, there is enough consensus on care benchmarks for early psychosis researchers to clearly define ‘appropriate care’ and precisely delimit youths’ pathways thereto. Also, most early intervention programs for psychosis target age groups that match our review’s age-based selection criterion.

45There is an evidence base for the adequacy of treatment for mental disorders other than psychosis. Efforts to quantify treatment delays have also expanded to more disorders, with the adoption of DUI measures in bipolar, anxiety and mood disorders. Yet, specific inquiries into pathways to care across these disorders, at least with respect to youth-focused literature, remain limited, as does our understanding of the association between pathways to care and treatment delays.

46The concept of appropriateness of pathway contacts warrants reflection. In early psychosis, contacts following the onset of frank psychotic symptoms that do not result in the commencement of psychosis-specific treatment can be viewed as missed opportunities for early intervention and prevention. More generally in youth, however, mental health symptom presentations are often transient and overlapping, and sometimes difficult to distinguish from developmentally normative behavioural or mood changes. It may therefore be difficult to establish an optimal ‘pathway to care’ in the broad field of youth mental health, and especially challenging to determine whether and when individuals reach an appropriate service. Two identical pathways may, in one case, reflect the appropriate use of a stepped-care model or, in another case, an inappropriately complex pathway. Moreover, even for similar problems, different individuals may have different optimal endpoints, based on available services, individual preferences, previous experiences, etc. Such complexities notwithstanding, studies on pathways to care can yield a greater understanding of how treatment gets delayed; and help identify the key agents involved in young peoples’ help-seeking processes and targets for outreach.

47It has been argued that ‘one-stop’ multidisciplinary integrated youth services can improve pathways to mental healthcare for young people. A central tenet of these services is the concept that ‘every door is the right door’. Such services aim to cater to youths with a range of needs (e.g., physical health, sexual health, mental health, housing, etc.) and types/severities of mental health problems. Examples of integrated youth services initiatives includes headspace in Australia, Jigsaw in Ireland, Youthspace in Birmingham, UK and ACCESS Open Minds, Foundry and Youth Wellness Hubs Ontario in Canada.

48Only one study in our review focused on pathways to care at a cross-diagnostic service that addressed severe and complex mental health conditions. We strongly recommend that the transformation of youth mental healthcare, including the establishment of youth hubs within community settings, be accompanied by increasing study of pathways to this presumably desirable endpoint. Such research is pertinent given young people’s preferences for community-based settings for mental healthcare.

49Pathways to care are quite variable across geographies, reflecting differences in healthcare, social, and cultural contexts. Many studies reported the attributes of their healthcare systems that may have influenced pathways to care. Importantly, individuals contacted many providers before reaching even those services that had open referral systems. This is perhaps unsurprising, given that, at least in psychosis, service configuration alone does not appear to impact treatment delays. This finding underscores the importance of early identification and outreach in reducing treatment delays as rapid access to care depends not only on systemic factors, but also on such influencers of help-seeking such as stigma, mental health literacy, and awareness of available services.

50Notably too, some studies reported longer referral delays than help-seeking delays, suggesting that the delay in treatment was attributable more to the care system itself. One can therefore conclude that the effort to reduce treatment delays and simplify pathways has to be directed at both the help-seeking and the referral components of treatment delay.

51The importance of primary care physicians prevailed in settings promoting ‘stepped care’ or general practitioner-gatekeeper models (e.g., Canada, Australia and Western Europe). Some contexts that allowed direct access to specialized care were likely to report self- or family-initiated referrals. In general, the role of general practitioners seems to be influenced by features of the healthcare system such as the availability and affordability of private or public mental health professionals.

52Our review included studies from both low- and middle-income countries (LMICs) and high-income countries. With more than 80% of the world’s population, LMICs deploy less than 20% of the world’s mental health resources. Often in LMICs, specialized care is inaccessible to many. These differences were reflected in our review. Certain LMIC-based studies described a difficulty in accessing formal mental healthcare, and cultural factors that influenced help-seeking (e.g., faith healers). More pathways to care research is needed in LMICs that have begun emphasising the integration of youth mental healthcare into existing community structures such as school, primary care, and community campaigns. Such research can yield valuable insights on whether pathways to mental health care are simplified when addressed through larger public health promotion and development initiatives.

53Notably, only four studies were from the United States, a country that otherwise generates volumes of mental health research. This suggests that interest in pathways to care may itself be a feature of public healthcare systems.

54Studies on pathways to care need to better report on the organization of local mental health services/systems, and beliefs about illnesses and services. This would help contextualize the appropriateness of potential routes to care across contexts.

55Many challenges remain in the assessment of pathways to care. The lack of standardization in the measurement of pathways to care is a major limitation that, in psychosis research, has been identified for over a decade.

56Wide variance in the definitions of start- and endpoints of pathways; and what and who constitutes a help-seeking contact limits our ability to compare results across studies. In many cases, the instruments chosen to assess pathways to care had a major influence on findings. Studies varied in their inclusion of formal, informal and ‘novel’ (e.g., web-based) contacts. The only study that specifically probed it, found that the internet figured prominently in the help-seeking process.

57Only three studies mentioned being guided by a theoretical framework, despite the frameworks for help-seeking behaviour and service use being available since the early 1990s and having been modified for mental healthcare pathways research.

58Studies on pathways to care are often premised on assumptions about the desirability of fewer contacts and, less frequently, the undesirability of certain types of contacts. Most studies are descriptive and provide estimates of individual and aggregate numbers and types of contacts made before a defined endpoint. However, evidence is lacking for whether more contacts along the pathway necessarily translate into longer treatment delays. Factors other than simply the number and type of contacts (e.g., waitlists, multiple encounters with the same contact, multiple contacts ending in evaluation but no treatment, etc.) may have a greater impact on treatment delays. Furthermore, reports of the numbers and types of help-seeking contacts do not reveal whether different services were accessed concurrently; whether appropriate treatments or referrals were offered and declined; or whether contacts met the individual’s needs. Also, notably absent is any measurement of how youths themselves perceived various help-seeking contacts.

59To advance research on pathways to youth mental healthcare and, thereby, youth mental health outcomes, we outline some key recommendations informed by our review. An important first step is standardization in the reporting of pathways to care. Specific recommendations in this regard are:

60Making it a standard to use and report theoretical frameworks in pathways to care research would facilitate better comparability across studies, more meaningful syntheses of extant knowledge, and easier identification of gaps.Studies on pathways to care should define pathways clearly, specifying start and endpoints.Studies should describe their intended methods of assessing pathways to care, justifying the choice of methodology in relation to study aims and the chosen theoretical framework. Ideally, an instrument with established psychometric properties should be used. Where a novel instrument is used, its psychometric properties must be established and/or described.The instruments should use a clearly specified timeframe, and techniques such as anchor dates should be employed to reduce the effects of telescoping bias, whereby events are recalled as occurring earlier or more recently than they actually did. This will allow for the accurate estimation of treatment delay indices.Studies should report on whether specific types of contacts were defined a priori or post hoc after collecting personal narratives, and whether specific types of contacts such as informal contacts (e.g., friends) and online resources were probed for in the interview.Studies should describe key features of the healthcare context (e.g., universal healthcare, access based on insurance, etc.) and referral system (e.g., walk-in access; need for a referral from a general practitioner, etc.) of their study setting.

61Making it a standard to use and report theoretical frameworks in pathways to care research would facilitate better comparability across studies, more meaningful syntheses of extant knowledge, and easier identification of gaps.

62Studies on pathways to care should define pathways clearly, specifying start and endpoints.

63Studies should describe their intended methods of assessing pathways to care, justifying the choice of methodology in relation to study aims and the chosen theoretical framework. Ideally, an instrument with established psychometric properties should be used. Where a novel instrument is used, its psychometric properties must be established and/or described.

64The instruments should use a clearly specified timeframe, and techniques such as anchor dates should be employed to reduce the effects of telescoping bias, whereby events are recalled as occurring earlier or more recently than they actually did. This will allow for the accurate estimation of treatment delay indices.

65Studies should report on whether specific types of contacts were defined a priori or post hoc after collecting personal narratives, and whether specific types of contacts such as informal contacts (e.g., friends) and online resources were probed for in the interview.

66Studies should describe key features of the healthcare context (e.g., universal healthcare, access based on insurance, etc.) and referral system (e.g., walk-in access; need for a referral from a general practitioner, etc.) of their study setting.

67The emergence of integrated youth services that, across geographic contexts, strive to adhere to common principles provides both a framework and an impetus for standardising the measurement of pathways to care. In addition to addressing the considerations for reporting of pathways to care outlined above, a standardized measure for pathways to care to be used across youth services should be relevant to and feasible for implementation in a range of contexts (urban, rural, Indigenous, high- or low-income, etc.). An ideal measure would capture pathways into the service (e.g., walk-in, referral, etc.); what was offered at the service (e.g., evaluation, short- or long-term treatment, crisis intervention, etc.); and pathways out of the service.

68Integrated youth services aim to offer well-publicised, rapidly accessible entry into a range of services and supports (not only those pertaining to mental health). The implicit assumption that such broad-spectrum services translate into more direct pathways and shorter delays to appropriate mental healthcare needs empirical testing. Some integrated youth services only offer interventions to those with mild to moderate mental and substance use concerns, referring more complex cases to external services. Future research therefore needs to examine whether such integrated youth services also succeed in simplifying pathways to care for youth with complex presentations.

69A foundational principle of current endeavours to transform youth mental healthcare has been a commitment to making services youth-oriented, and engaging youths in service design and evaluation. Consistent with this, the creation or deployment of any standardized measure of pathways to care should be conducted in partnership with youths and their families, and should pay due regard to youths’ perceptions of their pathways into care. Future studies would also do well to enquire about e-pathways to care, as youths are known to turn to the internet and social media in seeking mental health help.

70Our potential for comparisons across contexts and populations was limited by the lack of a standard methodology for ascertaining and reporting pathways to care. Our review’s scope was shaped by its inclusion of only quantitative studies that tend to focus on numbers and types of help-seeking contacts. Other significant aspects of the help-seeking process, such as beliefs about illnesses, and perceived barriers and facilitators to help-seeking, are largely found in qualitative analyses of pathways to care. Quantitative and qualitative approaches can have complementary potentialities in pathways to care research. Our age-based criterion was deliberately broad to accommodate studies that may have included, but not solely focused on, youths. However, this impedes our confidence in the applicability of our findings to exclusively youth-focused settings.

71Across contexts, young people’s pathways to mental healthcare are often complex and involve various formal and informal agents. Further research is necessary to better understand, and ultimately, to simplify and streamline pathways to appropriate services. This is an essential step towards ensuring easier, timelier access to care and, thereby, shaping youth mental health outcomes. More research is needed to address critical gaps in our knowledge of young people’s pathways to care for problems other than psychosis; the determinants of pathways; and the help-seeking behaviours of and service responses to underserved groups such as Indigenous youths, youth in protection/welfare systems, and homeless youths.

72Below is the link to the electronic supplementary material.