Naomi ZumsteinpersName, Florian RiesepersName (2020date). “Defining Severe and Persistent Mental Illness-A Pragmatic Utility Concept Analysis.” Frontiers in Psychiatrytitle·j.
Object of study — definitions of severe and persistent mental illnessterm·dom (SPMIterm·dom) in the health literature
Research question — how SPMIterm·dom is defined, and whether it is clearly and consensually defined with clear characteristics, preconditions, outcomes and boundaries
Contribution — SPMIterm·dom is a partially mature conceptterm·mth; its definition may rest on diagnosisterm·dom, disabilityterm·dom and durationterm·dom, operationalized for each context
phenomenon · claim-type · cross-domain-transfer · mechanism · claim-relation · replication-status · limitation — grounded in the full text| phenomenon | the lack of a consensus definition of severe and persistent mental illness — “The concept of severe and persistent mental illness (SPMI) lacks a consensual definition.” | |
| claim-type | critique — “We argue that the search for a grand theory of SPMI is misdirected because it ignores that the aim of every individual research project informs the choice of theoretical frameworks and the context-dependence of the operationalization.” | |
| method — evidence | — “We chose a particular variation of a systematic review, namely, PU concept analysis, as developed by Morse et al. ( 17 – 19 ).” | |
| object of study — evidence | — “Our objective was to clarify the concept of SPMI and to explore the level of concept maturity through pragmatic utility (PU) concept analysis.” | |
| research question — evidence | — “Specifically, our research questions were: How is SPMI defined and what are the features of the definition, i.e., is SPMI clearly and consensually defined, fully described, with clear characteristics, demonstrated preconditions and outcomes, and clearly delineated boundaries?” | |
| contribution — evidence | — “Therefore, while an agreement on the principle three dimensions included in a definition may be possible (diagnosis, disability, and duration), their operationalization will have to be context-dependent and specific for the task at hand.” | |
| mechanism | definitions Variation between SPMI definitions comes mainly from different meanings of the concept's constituent features, i.e., which dimensions to include. — “Variations in definitions appeared to mainly originate from different meanings of the constituent features of the concept, i.e., which dimensions to include.” | mechanisms[0] |
| mechanism | consensus definition An international consensus definition may serve epidemiology but produce a clinically irrelevant category locally, because it neglects local conditions. — “Finally, while an international consensus definition may facilitate large scale epidemiological research, it may create a clinically irrelevant category on the local level as the definition neglects local conditions, specificities, and nuances and has, therefore, no practical use.” | mechanisms[1] |
| builds-on | Morse et al. — “We chose a particular variation of a systematic review, namely, PU concept analysis, as developed by Morse et al. ( 17 – 19 ).” | claim_relations[0] |
| builds-on | Bachrach — “This argument has already been put forward by Bachrach ( 53 ) but seemingly ignored in subsequent studies.” | claim_relations[1] |
| builds-on | Brigandt — “Following Brigandt ( 49 ), we believe that an essential function of concepts is to set a problem agenda.” | claim_relations[2] |
| sense | SPMI - kept as a heuristic device rather than rebutted — “Instead of rebutting the concept of SPMI as a whole due to the alleged problem of lack of consensus definition, we propose to keep it as a heuristic device.” | senses[0] |
| sense | lack of consensus - reframed from the United States' local and national level to the international level — “In subsequent papers, the authors highlighted the lack of consensus on the international level [e.g., ( 4 , 5 , 13 , 31 )]. Hence, the problem of “consensus definition” was reframed in a far-reaching way. Shifting the meaning of the notion “consensus” implies a different approach to find a consensus definition.” | senses[1] |
| sense | SPMI definition - the original NIMH definition as the starting point for locally useful definitions — “As a starting point for such locally useful definitions, we propose the original NIMH definition of SPMI (see results section) ( 6 ).” | senses[2] |
| sense | consensus definition - potentially detrimental, a one-size-fits-all approach — “Thus, a consensus definition might even be potentially detrimental to the target population, as it may lead to a “one size fits all” approach.” | senses[3] |
| relation | diagnosis —part-of→ NIMH definition — The original NIMH definition ( 6 ) comprised the three dimensions diagnosis, disability and duration — “The original NIMH definition ( 6 ) comprised the three dimensions diagnosis, disability and duration operationalized as follows:” — “The original NIMH definition ( 6 ) comprised the three dimensions diagnosis, disability and duration” | concept_edges[0] |
| weight | diagnosis —part-of→ NIMH definition — 0.85 — “The original NIMH definition ( 6 ) comprised the three dimensions diagnosis, disability and duration operationalized as follows:” | concept_edges[0] |
| relation | disability —part-of→ NIMH definition — The original NIMH definition ( 6 ) comprised the three dimensions diagnosis, disability and duration — “The original NIMH definition ( 6 ) comprised the three dimensions diagnosis, disability and duration operationalized as follows:” — “The original NIMH definition ( 6 ) comprised the three dimensions diagnosis, disability and duration” | concept_edges[1] |
| weight | disability —part-of→ NIMH definition — 0.85 — “The original NIMH definition ( 6 ) comprised the three dimensions diagnosis, disability and duration operationalized as follows:” | concept_edges[1] |
| relation | duration —part-of→ NIMH definition — The original NIMH definition ( 6 ) comprised the three dimensions diagnosis, disability and duration — “The original NIMH definition ( 6 ) comprised the three dimensions diagnosis, disability and duration operationalized as follows:” — “The original NIMH definition ( 6 ) comprised the three dimensions diagnosis, disability and duration” | concept_edges[2] |
| weight | duration —part-of→ NIMH definition — 0.85 — “The original NIMH definition ( 6 ) comprised the three dimensions diagnosis, disability and duration operationalized as follows:” | concept_edges[2] |
| relation | SPMI —distinct-from→ disease entity — SPMI is not a disease entity — “SPMI is not a disease entity but rather refers to a particular population.” — “SPMI is not a disease entity” | concept_edges[3] |
| weight | SPMI —distinct-from→ disease entity — 0.8 — “SPMI is not a disease entity but rather refers to a particular population.” | concept_edges[3] |
| replication | not-applicable | |
| no type for this | severe and persistent mental illness → chronic mental illness — the NIMH introduced the new phrase in 1987 to replace the older one, because chronicity connoted incurable illness — “The phrase “severe and persistent mental illness” was introduced by the NIMH in 1987 ( 6 ) to replace “chronic mental illness” (CMI) because of the association chronicity has had with continuous or incurable illness ( 3 , p. 10).” | unhoused_relations[0] |
| no type for this | SPMI → palliative psychiatry — SPMI is the target population of palliative psychiatry — “In other words, SPMI is considered the target population of palliative psychiatry, and palliative psychiatry attempts to promote person-centeredness and patient autonomy.” | unhoused_relations[1] |
| limitation | The iterative coding was mainly performed by one author. — “A significant limitation was that the iterative coding process was mainly performed by NZ.” | limitations[0] |
| limitation | The heterogeneity of dimensions prevented a structured assessment of preconditions, outcomes and boundaries. — “The heterogeneity of dimensions included in the definitions did not allow us to assess the preconditions, outcomes, and boundaries of the concept in a structured way.” | limitations[1] |
| limitation | The search was limited to English-language literature. — “The search was limited to English-only literature, and no year limits were applied.” | limitations[2] |
finding · quantity · variable · definition · key-term — grounded in the full text| finding | SPMI is only a partially mature concept, because definitions and terminology vary widely and its constituent features are not fully articulated. — “Therefore, we conclude that SPMI is a concept that is only partially mature because definitions and terminology vary widely across the literature and because the constituent features are—as of yet—not fully articulated and need further clarification.” | content_tags.findings[0] |
| finding | Half of the papers addressed the lack of a consensus definition. — “Half of the papers addressed the lack of a consensus definition ( 1 – 5 , 10 , 13 , 15 , 24 , 29 – 31 , 35 , 37 , 38 ).” | content_tags.findings[1] |
| finding | Nine articles discussed SPMI in connection with palliative care. — “Nine articles discussed SPMI in connection with palliative care ( 7 – 13 , 15 , 39 ).” | content_tags.findings[2] |
| finding | Among the palliative-care articles, only three address the lack of a consensus definition of SPMI. — “Among the palliative care-related articles, only three address the lack of a consensus definition of SPMI ( 10 , 13 , 15 ).” | content_tags.findings[3] |
| finding | The initialism SMI was used both for severe mental illness and for severe and persistent mental illness. — “For instance, the initialism SMI was not just used for “severe mental illness” but also for “severe and persistent mental illness” instead of SPMI [e.g., ( 5 , 24 , 29 )].” | content_tags.findings[4] |
| quantity | 440 — — “After removal of duplicates, the initial literature search identified 440 articles.” | content_tags.quantities[0] |
| quantity | 108 — — “After exclusion of articles based on title and abstract, 108 articles underwent full-text reading.” | content_tags.quantities[1] |
| quantity | 30 — — “The reference lists of these studies were checked for further resources, resulting in the inclusion of an additional two publications for a total of 30 articles.” | content_tags.quantities[2] |
| quantity | 222 — — “The operationalized definitions were applied to a stratified sample of 222 adult patients who were admitted to an inner-city community mental health center in the city of Philadelphia to estimate the prevalence of SPMI in this sample.” | content_tags.quantities[3] |
| quantity | 20 — — “Slade et al. ( 4 ) address the problem of prioritization in mental health care and surveyed current practice in England by obtaining written documentation from 20 agencies on the eligibility criteria they use for deciding whether someone should receive mental health care.” | content_tags.quantities[4] |
| quantity | Seventeen — — “Seventeen definitions of the severe and persistent mentally ill, which have been developed in the context of mental health care in the United States were reviewed.” | content_tags.quantities[5] |
| definition | dimension — “Dimension” refers to diagnosis, duration, and disability or, in this case, dysfunction.” | content_tags.definitions[0] |
| key-terms | severe and persistent mental illness, SPMI, PU concept analysis, concept maturity, palliative psychiatry, consensus definition, diagnosis, disability, duration, SIDDD dimensions, normal science |
| concept | severe and persistent mental illness — SPMI refers to a patient population rather than a disease entity — “Our findings suggest that SPMI is a partially mature concept that needs further clarification. We argue that the lack of a uniform definition is inherent to the problem: SPMI refers to a patient population rather than a disease entity, and the term has to be useful for different stakeholder purposes. Therefore, while an agreement on the principle three dimensions included in a definition may be possible (diagnosis, disability, and duration), their operationalization will have to be context-dependent and specific for the task at hand.” | concepts[0] |
| concept | SPMI — Severe and persistent mental illness — “Severe and persistent mental illness (SPMI) is associated with suffering in the affected persons and burden to their caregivers.” | concepts[1] |
| concept | NIMH definition — The term SPMI was introduced by a workgroup convened by the NIMH in 1987 ( 6 ) and defined SPMI as a function of the three “Ds”, namely, diagnosis, disability, and duration — “The term SPMI was introduced by a workgroup convened by the NIMH in 1987 ( 6 ) and defined SPMI as a function of the three “Ds”, namely, diagnosis, disability, and duration.” | concepts[2] |
| concept | diagnosis — included a variety of illnesses — “Diagnosis” included a variety of illnesses, “duration” referred either to the duration of symptoms, treatment or disability and “disability” was used for functional impairment or quality of life.” | concepts[3] |
| concept | duration — referred either to the duration of symptoms, treatment or disability — “Diagnosis” included a variety of illnesses, “duration” referred either to the duration of symptoms, treatment or disability and “disability” was used for functional impairment or quality of life.” | concepts[4] |
| concept | disability — was used for functional impairment or quality of life — “Diagnosis” included a variety of illnesses, “duration” referred either to the duration of symptoms, treatment or disability and “disability” was used for functional impairment or quality of life.” | concepts[5] |
| concept | PU concept analysis — explores the level of maturity of concepts by assessing their internal structure, use, representativeness, and/or relations to other concepts — “PU concept analysis explores the level of maturity of concepts by assessing their internal structure, use, representativeness, and/or relations to other concepts.” | concepts[6] |
| concept | mature concept — A concept is “mature” when it is well-defined, has clearly described characteristics, delineates boundaries, and documented preconditions and outcomes ( 18 ) — “The evaluation of maturity uses the following overlapping criteria: epistemological (clarity of definition, boundaries, attributes, preconditions, and outcomes), pragmatic (operationalization of the concept in research and practice), linguistic (use in various contexts), and logical (differentiation of the concept from other concepts when used in theory). A concept is “mature” when it is well-defined, has clearly described characteristics, delineates boundaries, and documented preconditions and outcomes ( 18 ). However, a concept is not required to meet all of these criteria.” | concepts[7] |
| concept | partially mature concept — concepts that are partially developed ( 17 , p. 88), for instance, when distinguishing features are not yet fully articulated and need further clarification — “Between the two extremes of immature and mature concepts are concepts that are partially developed ( 17 , p. 88), for instance, when distinguishing features are not yet fully articulated and need further clarification.” | concepts[8] |
| concept | partial maturity — According to MacLeod ( 50 ), the open-endedness (or partial maturity) of a concept refers to an as of yet not well-known aspect of reality — “According to MacLeod ( 50 ), the open-endedness (or partial maturity) of a concept refers to an as of yet not well-known aspect of reality.” | concepts[9] |
| concept | normal science — the default state of mature science, which is characterized by a broad consensus of the practitioners of a scientific field on fundamental questions ( 16 , p. 10) — “Consequently, there is a need to evaluate the maturity of the concept of SPMI for palliative psychiatry to become “normal science”, i.e., the default state of mature science, which is characterized by a broad consensus of the practitioners of a scientific field on fundamental questions ( 16 , p. 10).” | concepts[10] |
| concept | palliative psychiatry — not defined in this paper's own words; a proposed care paradigm whose target population is SPMI and which aims at “person-centeredness and patient autonomy” — “In other words, SPMI is considered the target population of palliative psychiatry, and palliative psychiatry attempts to promote person-centeredness and patient autonomy.” | concepts[11] |
| concept | SIDDD dimensions — safety, informal and formal support, diagnosis, disability, and duration — “The findings indicate that definitions of severe mental illness are based on the SIDDD dimensions, i.e., safety, informal and formal support, diagnosis, disability, and duration.” | concepts[12] |
| concept | three-dimensional definition — uses three criteria: diagnosis of psychosis, duration of service contact ≥ 2 years, and GAF ≤ 50 — “The “three-dimensional definition” uses three criteria: diagnosis of psychosis, duration of service contact ≥ 2 years, and GAF ≤ 50.” | concepts[13] |
| concept | two-dimensional definition — uses only the last two criteria, i.e., duration and dysfunction — “The “two-dimensional definition” uses only the last two criteria, i.e., duration and dysfunction.” | concepts[14] |
| concept | top-down — includes consultation of managers specifying principles for identifying priority groups, by applying official guidelines to the local level — “Top-down” includes consultation of managers specifying principles for identifying priority groups, by applying official guidelines to the local level.” | concepts[15] |
| concept | bottom-up — entails staff working together to amend these principles in the light of their experience with individual clients — “Bottom-up” entails staff working together to amend these principles in the light of their experience with individual clients.” | concepts[16] |
| concept | thought style — a commonsense idea, a “thought style” ( 51 ), which remained unquestioned and carried on from one paper to the other — “The lack of an international consensus definition seems to have become a commonsense idea, a “thought style” ( 51 ), which remained unquestioned and carried on from one paper to the other.” | concepts[17] |
| concept | reverse epistemology — the elaboration of the concept stands at the beginning of the explanation chain; it precedes the accumulation of empirical evidence — “Moreover, the call for a consensus definition of SPMI follows a kind of reverse epistemology as the elaboration of the concept stands at the beginning of the explanation chain; it precedes the accumulation of empirical evidence.” | concepts[18] |
| concept | useful definition — According to van der Steen ( 54 , p. 18), a useful definition in the life sciences must be clear, not circular, neither too broad nor too narrow, should not include accompanying features, and should refer to the features present (rather than referring to features that are absent) — “According to van der Steen ( 54 , p. 18), a useful definition in the life sciences must be clear, not circular, neither too broad nor too narrow, should not include accompanying features, and should refer to the features present (rather than referring to features that are absent).” | concepts[19] |
| concept | chronic mental illness — not defined in this paper's own words; the earlier label that the 1987 NIMH phrase was introduced to replace — “The phrase “severe and persistent mental illness” was introduced by the NIMH in 1987 ( 6 ) to replace “chronic mental illness” (CMI) because of the association chronicity has had with continuous or incurable illness ( 3 , p. 10).” | concepts[20] |
| concept | disease entity — not defined in this paper's own words; a disease category, which the authors deny SPMI is — “SPMI is not a disease entity but rather refers to a particular population.” | concepts[21] |
Full text — tagged
Defining Severe and Persistent Mental Illness-A Pragmatic Utility Concept Analysis Abstract The concept of severe and persistent mental illnessterm·dom (SPMIterm·dom) lacks a consensual definition. Variations in definitions stem above all from different meanings about the constituent features of the concept and how to operationalize them. Our objective was to clarify the concept of SPMIterm·dom and to explore the level of concept maturityterm·key through pragmatic utility (PU) concept analysis. Our findings suggest that SPMIterm·dom is a partially mature conceptterm·mth that needs further clarification. We argue that the lack of a uniform definition is inherent to the problem: SPMIterm·dom refers to a patient population rather than a disease entityterm·dom, and the term has to be useful for different stakeholder purposes. Therefore, while an agreement on the principle three dimensionsterm·def included in a definition may be possible (diagnosisterm·dom, disabilityterm·dom, and durationterm·dom), their operationalization will have to be context-dependent and specific for the task at hand.
Keywords: severe and persistent mental illnessterm·dom, SPMIterm·dom, palliative psychiatryterm·dom, pragmatic utility concept analysis, systematic reviewterm·mth Introduction Severe and persistent mental illnessterm·dom (SPMIterm·dom) is associated with suffering in the affected persons and burden to their caregivers. To provide optimal healthcare for these patients is a challenge. To advance service provision for this population, research is needed to guide treatments and resource allocation. However, several authors have highlighted that SPMIterm·dom lacks a consensual definition ( 1 – 5 ). Variations in definitions stem above all from different meanings about the constituent features of the concept and how to operationalize them. Consequently, researchers, policymakers, and healthcare providers may have various understandings of SPMIterm·dom, measure different phenomena, and deal with different patient groups. Therefore, an analysis of the definitional basis of SPMIterm·dom is needed to advance the field.
The term SPMIterm·dom was introduced by a workgroup convened by the NIMH in 1987 ( 6 ) and defined SPMIterm·dom as a function of the three “Ds”, namely, diagnosisterm·dom, disabilityterm·dom, and durationterm·dom. The workgroup aimed to improve service and policy planning in the United States ( 3 , pp. 13–14, 6 ). In the following years, several authors developed definitions of SPMIterm·dom to be used on the local level ( 1 – 4 ).
Recently, a debate has emerged that connects SPMIterm·dom with palliative care ( 7 – 15 ). A paradigm shift to “palliative psychiatryterm·dom” for SPMIterm·dom has been proposed as this population is at risk of therapeutic neglect and/or overly aggressive care within the existing care paradigms ( 13 ). However, this ongoing debate has so far neglected to scrutinize the closely linked concept of SPMIterm·dom. SPMIterm·dom plays a crucial epistemic role in the formation of a palliative psychiatryterm·dom paradigm as it refers to its target. The current debate about palliative psychiatryterm·dom marks a “pre-paradigm” stage as the description lacks conceptual consensus ( 16 , p. 47). The idea of palliative psychiatryterm·dom itself may remain immature as long as the level of conceptual maturity of SPMIterm·dom has not been assessed. Consequently, there is a need to evaluate the maturity of the concept of SPMIterm·dom for palliative psychiatryterm·dom to become “normal scienceterm·mth”, i.e., the default state of mature science, which is characterized by a broad consensus of the practitioners of a scientific field on fundamental questions ( 16 , p. 10).
Hence, our objective was to clarify the concept of SPMIterm·dom in the health sciences and to explore the level of concept maturityterm·key through a particular form of systematic literature review, Pragmatic Utility (PU) concept analysis ( 17 – 19 ). Specifically, our research questions were: How is SPMIterm·dom defined and what are the features of the definition, i.e., is SPMIterm·dom clearly and consensually defined, fully described, with clear characteristics, demonstrated preconditions and outcomes, and clearly delineated boundaries? The need for a consensus definitionterm·key of SPMIterm·dom has emerged in specific domains with particular ways of understanding and reasoning of the issue. Therefore, we also aimed to explore how the problem of defining SPMIterm·dom was initially framed and followed up by various authors.
Methods Literature Identification and Selection Following a scoping search, a systematic search of the following databases was carried out to guarantee adequate and efficient identification of pertinent research related to the topic of interest and to minimize publication bias ( 20 ): EMBASE, SCOPUS, PsychINFO, MEDLINE, and Web of Science. Search tools such as medical subject headings (MeSH), Boolean operators, truncations, and positional operators were used. The search was limited to English-only literature, and no year limits were applied. The search strategy was developed in consultation with a research librarian, and the search strings adapted to suit the search style of specific databases used. In all databases, single and combined search terms included the terms “severe”, “serious”, “chronic”, “mental”, “illness”, “disorder”, “disease”, as well as “definition”, “concept”, “theory”, “model”, “method”, “concept model”, “concept framework”, and “approach”. The scoping search identified the article by Schinnar et al. ( 2 ) as a key article on the topic of definitions of SPMIterm·dom. Subsequently, a forward search in Google Scholar was performed to identify further studies that may have been missed in the primary database searches for inclusion ( 21 , p. 121).
Only publications that develop, propose, or describe a definition, theory, model, approach, or framework (theoretical or empirical) for defining, assessing, analyzing, and/or reporting SPMIterm·dom were included for the concept analysis (inclusion criteria). The included studies could be both empirical (qualitative and/or quantitative methods and systematic reviews) or theoretical (theory formation/development based on literature reviews and/or experience). The studies which were identified through the database and citation searches were downloaded into a reference management database and deduplicated. All titles and abstracts were then screened against inclusion criteria. References not meeting the inclusion criteria were excluded. A random sample of 20 articles (10 classified as not relevant, 10 as relevant by the first reviewer, NZ) was drawn, and a blind classification was performed by the second reviewer, FR. Classification differed in only one case that was resolved by subsequent discussion between reviewers 1 and 2, with reviewer 1’s classification prevailing.
Organization and Structuring of the Literature Articles were coded using MAXQDA 12 to chart the data. The organization and structuring of the literature were initially undertaken by NZ and then reviewed by FR. Relevant articles were first organized according to general features such as publication year, origin, the field of application, type of studies, and focus of the publication. A more in-depth structuring of the papers was guided by the following analytical questions: Which definitions were used in the literature? Is there a consensus within the literature about the definition? Was the lack of a consensus definitionterm·key addressed and followed up in the literature? Which definitions of SPMIterm·dom were used in relationship with the topic of palliative care? Was the lack of a consensus definitionterm·key addressed and followed up in the literature about palliative care? The coding frame for this in-depth structuring was based on an iterative deductive and inductive category formation ( 22 , p. 93).
Concept Analysis We chose a particular variation of a systematic reviewterm·mth, namely, PU concept analysisterm·mth, as developed by Morse et al. ( 17 – 19 ). PU concept analysisterm·mth explores the level of maturity of concepts by assessing their internal structure, use, representativeness, and/or relations to other concepts. It serves, among other things, to determine the need to refine or clarify a concept or to examine the congruence between the definition of the concept and the way it has been operationalized ( 17 , pp. 75–79). The evaluation of maturity uses the following overlapping criteria: epistemological (clarity of definition, boundaries, attributes, preconditions, and outcomes), pragmatic (operationalization of the concept in research and practice), linguistic (use in various contexts), and logical (differentiation of the concept from other concepts when used in theory). A concept is “mature” when it is well-defined, has clearly described characteristics, delineates boundaries, and documented preconditions and outcomes ( 18 ). However, a concept is not required to meet all of these criteria. Between the two extremes of immature and mature conceptsterm·mth are concepts that are partially developed ( 17 , p. 88), for instance, when distinguishing features are not yet fully articulated and need further clarification.
Rather than a series of steps, PU concept analysisterm·mth is a non-linear, iterative process. It contains, apart from the initial clarification of the purpose of the analysis the following guiding principles ( 23 ): a) identification and selection of relevant literature to ensure validity of the concept analysis, b) organization of literature in a general way, c) structuring literature through decontextualization to reveal general features of the concept, and d) formulation of key analytical questions to derive consistent dimensionsterm·def and boundaries of the concept. Also, we explored how the problem of defining the concept was framed and followed up by analyzing the underlying assumptions and premises.
Results Identified Literature After removal of duplicates, the initial literature search identified 440 articles. After exclusion of articles based on title and abstract, 108 articles underwent full-text reading. Based on the full-text, 28 articles were identified as relevant for our analysis. The reference lists of these studies were checked for further resources, resulting in the inclusion of an additional two publications for a total of 30 articles. For a flow chart of the literature identification and selection process, see Figure 1 .
Flow chart of the selection process.
The articles in our final sample included descriptive studies ( 1 , 3 , 5 , 8 , 10 , 24 – 32 ), systematic literature reviews ( 2 , 4 , 9 , 15 ), a non-systematic literature review ( 33 ), qualitative case studies ( 11 , 34 , 35 ), a mixed-method approach ( 36 ), and theoretical papers ( 7 , 12 , 13 , 37 – 40 ). Eleven articles originated in the United States, five in Canada, three each in Australia and Italy, two in Switzerland, one each in Sweden, Taiwan, Belgium, Netherlands, New Zealand, and the United Kingdom. Articles were distributed over time since the initial article in 1990 ( 2 ) (see Table 1 ). Definitions and concepts of SPMIterm·dom in included studies were based on psychiatry (eleven), nursing (six), psychology (three), social science (three), rural, public, population health (three), social work (three), and ethics (one). Findings from our review drawn from the collected articles resulted in the following themes in relation to which SPMIterm·dom is discussed: health services, care practices, palliative care, recovery, rehabilitation, social support, rehospitalization, and family adaptation. For characteristics of the included articles and the descriptions and definitions used, see Table 1 .
Summary table of included articles.
Definitions of Severe and Persistent Mental Illnessterm·dom The analysis of the descriptions and definitions used across the studies reveals an abundance and inconsistent use of different expressions, phrases, and initialisms. For instance, the initialism SMI was not just used for “severe mental illness” but also for “severe and persistent mental illnessterm·dom” instead of SPMIterm·dom [e.g., ( 5 , 24 , 29 )]. The initialism SPMIterm·dom was used for “severe and persistent mental illnessterm·dom” as well as “serious and persistent mental illness” ( 7 ). Half of the papers addressed the lack of a consensus definitionterm·key ( 1 – 5 , 10 , 13 , 15 , 24 , 29 – 31 , 35 , 37 , 38 ). Nine articles of those which addressed the lack of a consensus definitionterm·key also discussed which dimensionterm·def to include and how to operationalize them ( 1 – 5 , 24 , 29 , 31 , 38 ). Initially, the phrase “lack of consensus” referred to a lack of consensus among researchers, mental health planners, and policymakers on the local and national level in the United States ( 1 – 3 ). In subsequent papers, the authors highlighted the lack of consensus on the international level [e.g., ( 4 , 5 , 13 , 31 )]. Hence, the problem of “consensus definitionterm·key” was reframed in a far-reaching way. Shifting the meaning of the notion “consensus” implies a different approach to find a consensus definitionterm·key.
Variations in definitions appeared to mainly originate from different meanings of the constituent features of the concept, i.e., which dimensionsterm·def to include. Associations and forms of the dimensionsterm·def varied according to the different situations or conditions in which the concept was being used. Commonly, the definitions consisted of three dimensionsterm·def, namely diagnosisterm·dom, durationterm·dom, and disabilityterm·dom. “Diagnosisterm·dom” included a variety of illnesses, “durationterm·dom” referred either to the durationterm·dom of symptoms, treatment or disabilityterm·dom and “disabilityterm·dom” was used for functional impairment or quality of life. Slade et al. ( 4 ) proposed a five-dimensional framework for developing definitions at a local level. Ruggeri et al. ( 5 ) tested two operationalized definitions in two different settings, a narrow one (the “three-dimensional definitionterm·dom”) based on the National Institute of Mental Health definition ( 6 ) and a broad one (the “two-dimensional definitionterm·dom”).
Nine articles discussed SPMIterm·dom in connection with palliative care ( 7 – 13 , 15 , 39 ). Four out of the nine papers were theoretical ( 7 , 12 , 13 , 39 ) and five empirical ( 8 – 11 , 15 ). The definitions of SPMIterm·dom used in these papers were those from the National Institute of Mental Health ( 6 ), Schinnar et al. ( 2 ), Ruggeri et al. ( 5 ), and Woods et al. ( 15 ). Woods et al. refer to Ruggeri et al. ( 5 ) and the state definition of serious mental illness of Virginia (USA) ( 43 ). Among the palliative care-related articles, only three address the lack of a consensus definitionterm·key of SPMIterm·dom ( 10 , 13 , 15 ). However, none of them discussed and followed up on this issue in detail.
Milestone Publications We identified the definitions in the papers by the NIMH ( 6 ), Schinnar et al. ( 2 ), Slade et al. ( 4 ), and Ruggeri et al. ( 5 ) as the most impactful on subsequent publications and therefore present them in more detail here.
The phrase “severe and persistent mental illnessterm·dom” was introduced by the NIMH in 1987 ( 6 ) to replace “chronic mental illnessterm·dom” (CMI) because of the association chronicity has had with continuous or incurable illness ( 3 , p. 10). The task of the workgroup convened by the NIMH in 1987 was to develop a national definition of “severe and persistent mental illnessterm·dom” and builds on the work of Goldman et al. ( 42 ), who discussed how to define and count the “chronically mentally ill” (CMI). The objective of the NIMH workgroup was to reduce the variance in counts of persons affected by SPMIterm·dom to improve service and policy planning and to increase the congruence between federal and state guidelines for beneficiaries of federal and state economic and social support programs ( 3 , pp. 13–14). Also, the workgroup proposed to operationalize the definition using measures of disabilityterm·dom or dysfunction that are congruent with the Social Security Administration’s approach to defining disabilityterm·dom. However, the workgroup also suggested that specific thresholds on the measures could vary for state and local purposes.
The original NIMH definitionterm·dom ( 6 ) comprised the three dimensionsterm·def diagnosisterm·dom, disabilityterm·dom and durationterm·dom operationalized as follows: “Diagnosisterm·dom: A major mental disorder according to DSM-III-R: a major affective, non-organic psychotic disorder or a disorder that may lead to a chronic disabilityterm·dom such as a borderline personality disorder. Disabilityterm·dom: Severe recurrent disabilityterm·dom resulting from mental illness. The disabilityterm·dom results in functional limitations in major life activities. Individuals must meet at least two of the following criteria on a continuing or intermittent basis: (1) Is unemployed, is employed in a sheltered setting or supportive work situation, or has markedly limited skills and a poor work history (2) Requires public financial assistance from out-of-hospital maintenance and may be unable to procure such assistance without help (3) Has difficulty in establishing or maintaining a personal social support system (4) Requires help in basic livings skills such as hygiene, food preparation, or money management (5) Exhibits inappropriate social behavior which results in intervention by the mental and/or judicial system; Durationterm·dom: Treatment history meets one or both of the following criteria: (1) Has undergone psychiatric treatment more intensive than outpatient care more than once in a lifetime (e.g., crisis response services, alternative home care, partial hospitalization, or inpatient hospitalization) (2) Has experienced an episode of continuous, supportive residential care, other than hospitalization, for a period long enough to have significantly disrupted the normal living situation.”
Schinnar and his colleagues ( 2 ) used the NIMH definitionterm·dom as a reference point because it reflects a national perspective on SPMIterm·dom. The authors aimed to define and operationalize “severe and persistent mental illnessterm·dom” to obtain better national and local prevalence rates of SPMIterm·dom and to assess program effectiveness for recovery from SPMIterm·dom. Seventeen definitions of the severe and persistent mentally ill, which have been developed in the context of mental health care in the United States were reviewed. Based on the narrative descriptions of SPMIterm·dom, each criterion was operationalized. The operationalized definitions were applied to a stratified sample of 222 adult patients who were admitted to an inner-city community mental health center in the city of Philadelphia to estimate the prevalence of SPMIterm·dom in this sample. Their findings suggest that there is general agreement in the literature that diagnosisterm·dom, disabilityterm·dom, and durationterm·dom criteria in some form are necessary to define serious mental illness. The authors state that there is uncertainty about the relevant diagnostic categories, the nature and degree of disabilityterm·dom, the length of illness, and the relative importance of each. According to the authors, there is a general consensus that the durationterm·dom criterion should reflect the persistence of disabilityterm·dom and not the durationterm·dom of illness or treatment time. The authors posit that departures from the consensus definitionterm·key, which uses persistence of disabilityterm·dom as a criterion for the durationterm·dom, will and should occur within specific communities with special needs.
In a subsequent study ( 1 ), the authors highlight that mental health care needs are too heterogeneous across the United States, and resources available to meet such needs are too unevenly distributed to make a consensus definitionterm·key of SPMIterm·dom a practical tool. Hence, federal agencies should promote the development of a consistent framework for counting and reporting services for persons affected by SPMIterm·dom. The authors underline that states should remain free to develop definitions that suit their local needs best. However, they should be encouraged to develop definitions in a manner that is consistent with the components of the framework.
Slade et al. ( 4 ) address the problem of prioritization in mental health care and surveyed current practice in England by obtaining written documentation from 20 agencies on the eligibility criteria they use for deciding whether someone should receive mental health care. The study also included surveys of government departments, user groups, and professional bodies. The authors propose to include a “top-downterm·mth” and “bottom-upterm·mth” process in formulating a definition. “Top-downterm·mth” includes consultation of managers specifying principles for identifying priority groups, by applying official guidelines to the local level. “Bottom-upterm·mth” entails staff working together to amend these principles in the light of their experience with individual clients. The findings indicate that definitions of severe mental illness are based on the SIDDD dimensionsterm·dom, i.e., safety, informal and formal support, diagnosisterm·dom, disabilityterm·dom, and durationterm·dom. This particular set of components can be used as a framework for definitions at a local level.
Ruggeri and her colleagues ( 5 ) tested two operationalized definitions of “severe mental illness” (SMI) to calculate prevalence rates of SMI in two catchment areas in Europe (South London and South Verona). The prevalence rates were calculated according to a narrow (three-dimensional), and broad (two-dimensional) operationalized definition of SMI, which were derived from the NIMH ( 6 ) definition. The “three-dimensional definitionterm·dom” uses three criteria: diagnosisterm·dom of psychosis, durationterm·dom of service contact ≥ 2 years, and GAF ≤ 50. The “two-dimensional definitionterm·dom” uses only the last two criteria, i.e., durationterm·dom and dysfunction. Describing the definition as either “two-dimensional” or “three-dimensional” is, however, imprecise. “Dimensionterm·def” refers to diagnosisterm·dom, durationterm·dom, and disabilityterm·dom or, in this case, dysfunction. According to the authors, the two-dimensional definitionterm·dom is based on durationterm·dom and dysfunction only and applied to patients with and without psychotic disorders. Hence, the dimensionterm·def “diagnosisterm·dom” has actually not been excluded but rather operationalized differently.
Concept Maturityterm·key A concept is “mature” when it is well-defined, has clearly described characteristics, delineates boundaries, and documented preconditions and outcomes ( 18 ). The heterogeneity of dimensionsterm·def included in the definitions did not allow us to assess the preconditions, outcomes, and boundaries of the concept in a structured way. Therefore, we conclude that SPMIterm·dom is a concept that is only partially mature because definitions and terminology vary widely across the literature and because the constituent features are—as of yet—not fully articulated and need further clarification.
Discussion Research and improvement of service provision for SPMIterm·dom are hampered by a lack of a clear theoretical framework and, consequently, consensus what SPMIterm·dom actually is. Our aim was, therefore, to systematically review the existing literature on SPMIterm·dom for its definitions and to perform a PU concept analysisterm·mth to examine the concept’s maturity. We found inconsistent use of terminology and perpetual confusion about the constituent features of the concept and, consequently, different operationalizations of the dimensionsterm·def. We conclude that SPMIterm·dom is—as of yet—a partially mature conceptterm·mth that requires further theoretical development to become useful. Following Brigandt ( 49 ), we believe that an essential function of concepts is to set a problem agenda. Also, partial maturityterm·mth is not a hindrance to practical research. According to MacLeod ( 50 ), the open-endedness (or partial maturityterm·mth) of a concept refers to an as of yet not well-known aspect of reality. However, a concept is not required to give a correct representation of it to be scientifically useful.
The Lack of a Consensus Definitionterm·key Several studies have addressed the problem of how to define and operationalize the concept ( 1 – 5 , 29 , 38 ). Since the seminal studies of Schinnar and his colleagues ( 1 – 3 ), only Slade at al. ( 4 ) as well as Ruggeri and her colleagues suggested how to operationalize the concept ( 5 , 29 ). Definitions used in articles about palliative care and SPMIterm·dom refer to four sources ( 2 , 5 , 6 , 15 ), and the lack of a consensus definitionterm·key of SPMIterm·dom is left commonly unaddressed.
Initially, the phrase “lack of consensus” referred to a lack of consensus among researchers, mental health planners, and policymakers on the local and national level in the United States ( 1 – 3 ). The main goal of finding a consensus definitionterm·key of SPMIterm·dom was to increase the congruence between federal and state guidelines for beneficiaries of federal and state economic and social support programs. For instance, the dimensionterm·def “disabilityterm·dom” was formulated in congruence with the Social Security Administration’s approach to defining disabilityterm·dom. The Social Security Administration is a national agency of the U.S., and its regulations on disabilityterm·dom may not be suitable as a reference point to develop a definition that will be applied in other contexts. Moreover, definitions developed by institutions are commonly not based on empirical findings and not intended for scientific use but for management issues.
Interestingly, these statements about the scale of the definition have been ignored in subsequent studies as the problem of defining “severe and persistent mental illnessterm·dom” (SPMIterm·dom) or “severe mental illness” (SMI) has been mainly framed as the lack of an international consensus definitionterm·key. Hence, the problem of consensus definitionterm·key was reframed in a far-reaching way as it implies a different approach to solving the problem of lack of consensus. The lack of an international consensus definitionterm·key seems to have become a commonsense idea, a “thought styleterm·dom” ( 51 ), which remained unquestioned and carried on from one paper to the other. The underlying assumption of this kind of problematization is that there should be a general theory of SPMIterm·dom.
We argue that the search for a grand theory of SPMIterm·dom is misdirected because it ignores that the aim of every individual research project informs the choice of theoretical frameworks and the context-dependence of the operationalization. SPMIterm·dom is not a disease entityterm·dom but rather refers to a particular population. Also, the claim for a consensus definitionterm·key implies a homogeneity of the target population, but all descriptions of the target population are, of necessity, combinations of medical (e.g., diagnostic assessment) and social (e.g., community functioning assessment) ( 52 ) as well legal criteria. Hence, the call for a standardized international definition of SPMIterm·dom is based on an incomplete understanding of the requirements that such a uniform definition would have to fulfill. Moreover, the call for a consensus definitionterm·key of SPMIterm·dom follows a kind of reverse epistemologyterm·mth as the elaboration of the concept stands at the beginning of the explanation chain; it precedes the accumulation of empirical evidence.
Finally, while an international consensus definitionterm·key may facilitate large scale epidemiological research, it may create a clinically irrelevant category on the local level as the definition neglects local conditions, specificities, and nuances and has, therefore, no practical use. SPMIterm·dom affects people who differ considerably with regards to their diagnoses, treatment histories, functional levels, and needs. Thus, a consensus definitionterm·key might even be potentially detrimental to the target population, as it may lead to a “one size fits all” approach. This argument has already been put forward by Bachrach ( 53 ) but seemingly ignored in subsequent studies. Although she refers to CMI, we believe that her argument also applies to “severe and persistent mental illnessterm·dom”.
Reaching a Useful Definitionterm·mth Instead of rebutting the concept of SPMIterm·dom as a whole due to the alleged problem of lack of consensus definitionterm·key, we propose to keep it as a heuristic device. Still, we argue against a consensus definitionterm·key for use on the international level. We argue that the search for a consensus definitionterm·key of SPMIterm·dom based on a single general theory is misguided as it ignores the context-dependence of health and healthcare. Future research on SPMIterm·dom ought to focus first on reducing ambiguity in the terminology and achieving context-dependent clarity about the meaning of the concept ( 54 , p. 10). Hence, to clarify the connection between the concept of SPMIterm·dom and its theoretical or empirical statements, there is a need to develop a specific context-dependent theoretical framework and to formulate a definition of SPMIterm·dom for use on the local level and to operationalize the constituent dimensionsterm·def according to specificities and needs in a given context.
As a starting point for such locally useful definitionsterm·mth, we propose the original NIMH definitionterm·dom of SPMIterm·dom (see results section) ( 6 ). According to van der Steen ( 54 , p. 18), a useful definitionterm·mth in the life sciences must be clear, not circular, neither too broad nor too narrow, should not include accompanying features, and should refer to the features present (rather than referring to features that are absent). Accordingly, we suggest the definition of SPMIterm·dom to be based on the three Ds, i.e., diagnosisterm·dom, durationterm·dom, and disabilityterm·dom. As constituent characteristics, they are abstract enough to define the concept regardless of the context in which it appears, yet unique enough to define and differentiate it from other concepts [cf. ( 18 , p. 388)]. Also, this allows a notion of SPMIterm·dom that is wide enough for quite different further developments. Given the context-dependence of the different constituent features of the definition, we argue that the constituent dimensionsterm·def then have to be further operationalized according to the priorities, needs, and subtleties prevailing in a given context. Local operationalizations would undoubtedly require an update of the diagnostic system in use (i.e., DSM 5 or ICD-11 instead of DSM-III-R) and would likely involve more subjective parameters in the disabilityterm·dom and durationterm·dom dimensionsterm·def (emphasizing SPMIterm·dom as illness concept rather than as healthcare planning construct).
SPMIterm·dom and Palliative Psychiatryterm·dom It has been postulated that a palliative approach in psychiatry has the potential to improve the quality of care, person-centeredness, and autonomy for patients affected by SPMIterm·dom ( 13 ). In other words, SPMIterm·dom is considered the target population of palliative psychiatryterm·dom, and palliative psychiatryterm·dom attempts to promote person-centeredness and patient autonomy. As to be expected from our concept analysis, a recent article on a novel therapeutic approach to SPMIterm·dom had to further specify its intended target population as “most serious forms of SPMIterm·dom, for instance, when the quality of life is seriously compromised” ( 55 ). We ultimately believe that a definition of SPMIterm·dom true to a palliative ethos of care will necessarily have to take the perspectives of the affected persons into account. Thus, a definition of SPMIterm·dom for use in the field of palliative psychiatryterm·dom should be negotiated among user groups and their caregivers, government agencies, insurers, and treatment teams who establish a set of principles to help guide the development of a definition. To a certain extent, this may allow for the self-assignment of persons affected by SPMIterm·dom to receive palliative care in psychiatry based on patients’ needs, goals, and preferences. However, the theoretical framework, as well as the empirical evidence for palliative psychiatryterm·dom, have to be developed further to be able to define its target population.
Hence, we propose a participatory/collaborative approach as a research strategy which combines theoretical and practical knowledge to generate new knowledge about a phenomenon. The underlying ethos of such a research strategy has been formulated by Burns ( 56 ). The work of Vreman et al. ( 57 ) illustrates how semi-structured group discussions and open session workshops among various stakeholders provided insights about the concept of unmet medical need (UMN).
Strengths and Limitations This is the first PU concept analysisterm·mth to evaluate the maturity of the concept of SPMIterm·dom. Important strengths of our systematic reviewterm·mth include a well-defined research question, an exhaustive search strategy, well-defined inclusion and exclusion criteria, and comprehensive data coding. A significant limitation was that the iterative coding process was mainly performed by NZ. However, the analysis of the coded data was conducted in close cooperation with both authors.
Conclusions A systematic concept evaluation is a fundamental step in the research process that should precede more formal research procedures such as operationalization or identification of the variables ( 18 ). It allowed us to identify ambiguities and inconsistency of constituent features of the concept of SPMIterm·dom. According to our PU concept analysisterm·mth, SPMIterm·dom refers to a partially mature conceptterm·mth that lacks a standard definition. We argue that this lack of a uniform definition is inherent to the problem: SPMIterm·dom refers to a patient population rather than a disease entityterm·dom, and the term has to be useful for different stakeholder purposes. Hence, the next step toward concept clarification and development consists of an assessment of the different perceptions of various stakeholders through a participatory oriented approach between researchers, persons affected by SPMIterm·dom, and multiple practitioners. Insights gained from semi-structured group discussions and open session workshops among these different stakeholders have then to be incorporated in a modified definition of SPMIterm·dom.
Data Availability Statement All datasets generated for this study are included in the article/supplementary material.
Author Contributions All authors contributed to the article and approved the submitted version.
Funding Open access publication was funded by the University of Zurich.
Conflict of Interest The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Acknowledgments The authors thank Jill Boruff, MLIS, Liaison Librarian to the School of Physical and Occupational Therapy, the School of Communication Sciences and Disorders, and the Department of Psychiatry, McGill University, Montreal, Canada References Associated Data Data Availability Statement All datasets generated for this study are included in the article/supplementary material.