 |
 |
|
|
|
Continuity of care (COC) is widely regarded as central to the provision of mental health services. It has recently assumed greater importance in conjunction with the decentralization of service delivery in mental health systems across Europe (1) and North America (2–4). Clinicians commonly assume that, if COC is absent from the services provided to patients with chronic and debilitating conditions, the result may be social isolation, economic hardship, and threats to quality of life. The concept has been characterized by some as a “strategic first choice” in the case of service planning and delivery for patients with severe and persistent mental illness (SPMI; 5). Theoretical statements regarding COC (for example, 6) have emphasized multiple component dimensions, reflecting its status as a theoretical construct, but as yet, there are no definitive statements on which dimensions are critical. In this study, we systematically examined the COC construct with the intent of identifying those critical dimensions. There are findings indicating that adequate COC is related to better symptom control and decreased length of hospitalization, and evidence suggests that it may be associated with improved mental health, functioning, and quality of life, as well as better medication compliance and decreased emergency room visits and rehospitalizations (3,7–10). Some authors assert that COC also makes care more cost-efficient (3). At the same time, many commentators have noted that COC is a conceptually underdeveloped, vague, and over-inclusive construct lacking a solid empirical foundation (3,6,8–12). For example, Johnson and others comment that “research on the measurement of continuity has been notably sporadic and discontinuous . . . reflected in the wide variety of definitions and operationalisations (3, p 140).” This situation—minimal but provocative evidence that continuity is associated with positive outcomes, together with lack of agreement on definitional and dimensional issues—suggests that it is timely to develop a more accurate definition of the COC construct. In turn, a consensual definition capable of operational formulations will prove useful in the development of valid empirical measures of COC for mental health services research. One aspect of the definitional problem that has not received extensive attention is that assessment of COC is almost universally conceptualized from the service provider’s point of view. Several operational measures that reflect this perspective have been proposed. Examples include “the proportion of patients transferred to another hospital during admission” (13); “the proportion of patients discharged from hospital who are referred to community services and remain in services 6 months after discharge” (14); and “the number of times the patient’s contact with care is terminated during the study period” (15). Although important from the perspective of mental health service administration, these measures fail to convey anything about the patient’s experience of COC, although patients and their families may be the only parties actually in a position to judge continuity across multiple service providers and settings. Consequently, there is rarely evidence to indicate whether patients actually experience the interventions and administrative policy changes designed to facilitate COC as improvments in service delivery. As noted, the extant mental health literature provides 1 source of information on the COC construct, that is, the perspective of clinicians, service administrators, or researchers studying mental health services. Clearly, a relevant but neglected source of information is the group of individuals with SPMI for whom mental health services are provided. This study involved the collection, qualitative analysis, and formulation of a pool of potentially relevant COC characteristics. Because of the numerous available conceptual and empirical definitions of COC, we did not assume that we knew its relevant attributes and dimensions a priori. Instead, we adopted an inductive approach: we systematically reviewed the relevant theoretical and empirical literature, and we conducted interviews with patients receiving mental health services and with their families. Methods and ResultsLiterature Review In each database, we selected search terms (specifically, subject headings and [or] keywords and [or] key word phrases) that included some aspect of COC alone or in combination. We examined the scope note for each search term for content and usage. We then examined a sample of 10 to 20 citations and abstracts generated by each search term to confirm its content and usage and to generate other related terms. To capture older articles on COC, we also included previously used search terms (as indicated in the scope notes). For example, the scope note for “continuity of patient care” indicates that, from 1968 to 1978, the topic was indexed under “comprehensive health care,” and from 1979 to 1990, it was indexed under “primary health care.” We therefore employed all 3 terms in the search strategy. The search of each database proceeded through 5 stages, each encompassing specified literature: 1) continuity of care, 2) medical (nonpsychiatric) conditions–medical care, 3) psychiatric disorders–mental health care, 4) the combination of medical care and COC, and 5) the combination of psychiatric–mental health care and COC. The main search results were validated by an independent search using a similar strategy and conducted in a different setting. Both searches were conducted by trained librarians (see Note). Most of the abstracts identified by the independent search (97%) had also been captured by the main literature search. The small proportion of abstracts (3%) that were judged to be on topic but missed by the main search suggested that the bulk of the appropriate literature on COC for SPMI patients had been identified. The literature search yielded a total of 1673 citations. When duplicate citations were eliminated, 1281 articles remained. Abstracts were then sorted into the categories of COC from the perspective of mental health services (n = 360) and COC from the perspective of general health services (n = 921). Next, 3 authors reviewed the 360 abstracts drawn from the mental health area for inclusion in the subsequent qualitative analysis. Abstracts were included if there was an explicit reference to COC, content associated with elaboration of the construct, or a reference to efforts to put into practice or empirically study the concept. Abstracts were excluded if the article was only available in a language other than English and (or) if the article dealt solely with financial considerations of service delivery in a managed care environment. In a randomly selected subgroup of 30 reviewed abstracts, the average pairwise agreement regarding inclusion was 76% (range 67% to 83%). This preliminary review led to the exclusion of a further 55 abstracts, yielding a final set of 305 articles. To initiate a more systematic review, 2 judges were first trained to extract statements that informally referred to COC, that reflected “theoretical and operational definitions of COC,” and (or) that provided “descriptions of techniques for promoting COC.” The judges and a coinvestigator first worked as a group to extract statements from 3 randomly chosen articles. Subsequently, the 2 judges worked independently on 10 randomly selected articles and resolved discrepancies through discussion. When the judges’ extractions were considered to be adequately consistent, each proceeded to an independent review of all 305 articles. The statements extracted by both judges were transcribed verbatim onto index cards. An example of a theoretical definition is “Each patient must be provided with an opportunity to establish a dependable and continuing relationship with a caregiver or caregivers who can help the patient navigate through the system of care” (2). An example of an operational definition is “the time interval between hospital discharge and resumption of community-based treatment” (16). An example of a technique for promoting COC is “Inpatient staff accompany patients to see their outpatient clinician prior to discharge” (17). This extraction process generated a total of 906 COC “attributes,” an average of 3.02 attributes per article. An informal review of the first 50 articles from the pool of 305 readily identified 11 recurring themes. For example, at the level of the service system, the importance of linkages between services, information availability, and continuity of place were frequently mentioned. Regarding treatment provision, critical dimensions of COC included comprehen- siveness, accessibility, longitudinality (that is, sustained care over time), emphasis on adherence or compliance, regularity of contacts, and attention to individual patient needs. In terms of the interaction between individual provider and patient, provider knowledge and the qualitative aspects of the therapy relationship emerged as critical to COC. This cursory review suggested that a more systematic approach to the literature would yield detailed information on the critical dimensions of the COC construct. Classification of Attributes. About one-third of the selected attributes (n = 279, 30.8%) were eliminated from further analysis because they met either of the following exclusion criteria: 1) the statement referred to a specific programmatic approach without specifying how the approach promoted COC (for example, “case management may be viewed as a vehicle for implementing continuity in the care of mentally ill persons”; 18) ; or 2) the statement provided only nonspecific assertions of the value of COC (for example, “obstacles to accessibility must be overcome”; 19). At this point, we concluded the process of selecting attributes from the literature. The next phase of the qualitative inquiry aimed at classifying the remaining 627 attributes into distinct conceptual domains. Two teams of 2 investigators classified the attributes. The teams worked independently and used distinct approaches. We assumed that correspondence between the 2 resulting classifications would indicate convergent validity for our understanding of the COC construct. Team A, comprising 2 research psychologists, used several techniques from qualitative methodology to organize the classification task. First, they sorted attributes into piles that seemed to reflect distinct primary “domains” of COC (for example, 20). Thus a clear distinction could readily be made between COC attributes reflecting policies of treatment programs and those reflecting characteristics of the provider–patient relationship; in other words, these 2 sets of attributes represented distinct conceptual domains of the COC construct. Second, they performed taxonomic analysis on each of the primary domains of COC attributes (21). This involved reviewing attributes assigned to the major superordinate domains and constructing “subdomains,” that is, essential differentiating characteristics nested within the major domain. The process was continued with those attributes assigned to subdomains to derive “categories,” that is, specific, concrete, and observable characteristics of the subdomain. The grounded theory concept of theoretical saturation (22) was used to define subdomains and categories: attributes were classified into more specific topic areas until no new subdomains or categories could be identified. This approach to the classification process resulted in a hierarchical model of the COC construct. Figure 1 presents the 4 domains and 8 subdomains of COC attributes identified by Team A. The figure indicates the number and proportion of attributes assigned to each domain or subdomain. The largest domain includes 437 (69.7%) selected attributes and reflects aspects of the organization and coordination of services. It addresses the policies, structural features, and procedures that treatment systems and services require to achieve COC. This domain was labelled “Service Delivery.” Within this broad area, 3 attribute subdomains were identified, corresponding to different levels at which service organization and coordination could occur. Attributes in subdomain 1 (System Administration) reflect activities to integrate services across boundaries between hospital and community or between mental health and medical health programs (for example, “senior clinicians and administrators from the area and community support services met monthly,” 16). In contrast, attributes in subdomain 2 (Program Delivery) reflect activities to promote COC within the frame of hospital- or community-based programs (for example, “the same clinicians provide clients with core services for the duration of their treatment in the program,” 23). Finally, attributes in subdomain 3 (Provider Action) reflect activities associated with the individual provider’s delivery and documentation of mental health services (for example, “provide telephone consultation on a timely basis,” 24).  Further, each Service Delivery subdomain was seen to encompass 3 attribute categories. Figure 2 presents these subdomains and their associated categories, with the number and proportion of attributes assigned to each. For the System Administration and Program Delivery subdomains, the 3 categories are similar and reflect 1) policies addressing the functional operation of the system or program, 2) specified structural linkages between programs in the system or between services in a given program, and 3) either actual service activities occurring in the system or interventions offered by the program. The 3 categories of Provider Actions are represented by 1) policies for documenting service contacts, 2) actions required to ensure COC during hospital admissions and at discharge, and 3) actions required to ensure COC in the provision of services at the community level.  A second broad domain, Accessibility, includes 63 (10.1%) attributes addressing spatial and temporal conditions that facilitate COC. Subdomains address Temporal Access (for example, “Patients receive the care they need at the most appropriate time,” 25), Geographic Access (for example, “The patient should, as much as possible, remain in his own milieu,” 26), and specific Barriers to Access (for example, “The delivery of some program components on a regional basis, while providing others on a local basis, could present a barrier to continuity of care,” 27). The third domain, Relationship Base, includes 67 (10.7%) attributes and addresses the quality of the relationship between patients and service providers. This domain has 2 subdomains. One addresses patients’ Relationship to the System (for example, “Provide continuity of care and caregivers, thus allowing patients to build an institutional transference so that the institution and its workers indeed become a source of support and advice,” 28). The second subdomain addresses qualities of the Relationship to the Provider (for example, “Pinch hitting: when individual service providers step outside their prescribed roles to undertake tasks usually performed by somebody else,” 9). The final broad domain, Individualized Care, has 60 (9.6%) attributes and addresses the extent to which service providers are sensitive to the personal, social, and cultural circumstances of patients, in addition to their clinical needs (for example, “the patient must be encouraged to participate as a partner in the process of service planning,” 2). Team B consisted of an epidemiologist and a psychiatrist, both with extensive research and clinical experience involving SPMI patients. These investigators individually sorted the 627 attributes and then reached a consensus on themes consistently reflected in the literature. In contrast to Team A’s development of a hierarchical model, Team B simply listed themes ranked according to the frequency of references in the literature. Table 1 presents the list of themes and their definitions.
Qualities of the specific patient–provider relationship and individual care plan had thematic priority in this classification. This individualized focus was followed by themes concerning aspects of continuity at the system level, that is, the accessibility, integration, and comprehensiveness of services. The remaining themes concerned service program characteristics, including the degree of flexibility in community services, the efficiency of managing client transitions between hospital and community, the degree of emphasis on collaboration and communication between programs or providers, and program perspectives on long-term care. Figure 3 compares the classifications derived by Teams A and B. There was a high degree of correspondence between the classifications: each of the 10 themes identified by Team B could be paired with a subdomain or category from the hierarchical model developed by Team A. However, not all categories of Team A’s hierarchical classification had a corresponding theme in Team B’s list, specifically, the categories Regulatory Function and Types of Activities in the subdomain System Administration and the category Types of Activities in the subdomain Program Delivery. With regard to the subdomain Provider Actions, the 2 categories identified by Team A as reflecting service delivery by a treatment provider in either the hospital or community setting were captured by the single theme identified as Transition Management by Team B. Apart from these relatively minor differences, the classifications derived by the 2 investigator pairs showed a high level of consistency across attribute groupings, suggesting a good level of convergent validity across these 2 different taxonomic analyses.
The results of the initial informal review and the 2 teams’ attribute classifications were subsequently addressed in a consensus meeting of all project coinvestigators. The characteristics of each primary theme (according to the informal review and Team B’s classification) or domain (according to Team A’s classification) were evaluated in regard to their priority for the provision of clinical services to SPMI patients. Consensus was reached on 6 aspects of COC that were regarded as critical for the development of a quantitative index of the construct (highlighted by italics in Figure 3): the patient–provider relationship, the degree of individualized care, accessiblity, linkages between services, comprehensiveness of services, and flexibility of the service location. Team A’s hierarchical model was addressed again in a final step of the classification process. Attributes within each domain, subdomain, and category were ranked by the 2 team members according to how clearly they expressed the definition for that aspect of the hierarchical model (1 = poor, 2 = weak, 3 = adequate, 4 = very good, and 5 = excellent). An example of a poor attribute from the Relationship Base domain is “continuity of caregivers refers to the specific relationship between the patient and treatment providers” (29). An example of an excellent attribute from the Program Delivery subdomain is “being admitted to the same unit and having the same staff members being responsible for the client on a repeated admission” (30). Ultimately, 125 exemplary attributes (that is, those with a consensus ranking of 4 or 5) were identified and reviewed for redundancy and applicability to questionnaire format. Of these, 47 attributes were discarded because they were not amenable to assessment via patient self-report, leaving 78 COC attributes drawn from the literature that were representative of the 2 attribute classifications and appropriate for formulation as self-report questionnaire items. Patient Interviews We informed respondents about the study and asked them to consent to be interviewed, as required by the research ethics committees at each university and in each health region. We obtained a purposive sample of 36 respondents: 27 SPMI patients (13 men and 14 women) and 9 respondents (6 mothers, 2 sisters, and 1 male partner) related to a patient with SPMI. Procedure. Each semistructured interview involved 4 sets of questions probing the following topic areas: 1) history of care (for example, type of service involvement and duration); 2) experienced continuity of care (for example, perceptions of service coordination); 3) personal experiences of high or low COC (for example, description of a period when services were felt to be “smooth, well-connected, and linked”); and 4) evaluation of the 78 COC attributes derived from the literature review, with elicitation of additional concepts related to COC from the patient’s perspective. Each interview was audio- recorded and transcribed. Analyses of the interview transcripts were undertaken in consultation with a qualitative researcher experienced in grounded theory methods. Transcribed interviews were reformatted with line numbering and a large right margin. We completely reviewed each transcript before we extracted attribute phrases. In a second review, material describing the interviewee’s experience of COC was flagged in the margin. This material was extracted and condensed into a phrase or a sentence. Whenever possible, we employed respondents’ actual words. Each extracted phrase was then reviewed to ascertain whether it corresponded to attributes defined from the literature review or whether it represented a new COC attribute. Results. From the 36 transcripts, we extracted 574 phrases capturing patient and (or) family member comments on COC. We compared each phrase with the 78 attributes already derived from the literature review. If the interview phrase paraphrased an attribute extracted from the literature, it was checked off the master list. If the phrase could not be matched with the items derived from the literature, we formulated a new item to represent the expressed meaning. For most of the interview phrases (531, or 92.5%), one or more literature-derived items already existed. However, we identified 43 (7.5%) additional attributes reflecting unique patient and family member perspectives on COC (for example, “My appointments can be more often if I am not doing well,” “It’s very hard to get in to see some professionals,” and “I have felt ‘lost in the system’”). With the exception of service flexibility, these additional attributes commonly highlighted more specific elements of the 6 critical COC aspects identified at the consensus meeting of project coinvestigators. For example, regarding the patient–provider relationship, interviewees emphasized the caregiver’s understanding and atittude. In terms of individualized care, patients and families highlighted the importance of advocacy, personal support, and patient- specific treatment. In terms of accessibility, respondents underscored the importance of readily available help, access to housing, and barriers to service. With regard to the links and connections between services, transitional management and communication emerged as important elements for patients. In terms of the comprehensiveness of services, uneven quality or gaps in services were paramount concerns. Formulating Self-Report Questionnaire Items The preceding steps yielded a pool of 121 provisional items assessing COC (78 from the mental health literature and 43 from interviews with SPMI patients and their families). From this, we believed we could formulate our initial self-report measure of patient perceptions of COC. Determining the items to be used for questionnaire construction and initial psychometric testing involved 2 phases. First, all coinvestigators reviewed the 121 items for comprehension, redundancy, use of language, and amenability to scaling. Next, 23 mental health professionals working with the SPMI population and 18 SPMI patients (none of whom participated in the interview phase) rated the items. As well, the 41 respondents were asked to comment on the items’ wording, clarity, and comprehended meaning. The respondents were also asked to indicate whether each item was a good candidate for describing COC (yes or no) and to rank the items by picking their “top 10" out of the list. The following algorithm was developed to combine and weight respondent endorsements and preferences: [(#yes – #no) + frequency of top 10 judgements]. We used this formula to score each item from 0 to a maximum of 58. Clinicians and patients were highly consistent in terms of which items were endorsed as useful for describing COC. Pearson correlation of the proportions of each group endorsing a given item, across all items, was r119 = 0.42, P < 0.0001. For items that were finally selected from the pool of 121, the average proportion endorsement for clinicians was 81.0% (SD 11.83%, range 53.3% to 100.0%), and for patients, 83.0% (SD 10.50%, range 61.5% to 100.0%). Items receiving scores of 21 or greater on the weighting algorithm were retained (n = 40), with 7 additional items kept to ensure coverage of the attribute domains and subdomains identified by Team A. These pretest activities resulted in a pilot questionnaire of 47 items. We attempted to balance positive and negative wording and to maintain a consistent reading level across the items. In a final step, we prepared standard definitions for terms (for example, primary caregiver, provider, program, and services) used throughout the measure. Items assessing demographic and clinical characteristics (that is, age, sex, current living arrangements, housing status, education, employment status, diagnosis, age when first diagnosed, and number of psychiatric hospitalizations) were also added in a separate section to examine relations between these characteristics and scale scores. The final version of the draft measure had an average Flesch–Kincaid reading level of grade 5.6; we called it the Alberta Continuity of Services Scale-Mental Health (ACSS-MH). Table 2 lists the 47 draft items and sorts them into the 4 COC domains illustrated in Figure 1. A field trial and evaluation of the psychometric characteristics of the ACSS-MH was implemented as part of a larger study of continuity in mental health services provision; it will be the subject of an independent report.
DiscussionThis paper presents our qualitative examination of the COC construct. We used an inductive approach to identify key attributes of COC from a comprehensive search of the mental health literature and supplemented these “expert” conceptions with material from interviews with SPMI patients and their families. These sources allowed us to systematically describe key characteristics of the COC construct. The qualitative investigation resulted in a draft self-report questionnaire designed to assess SPMI patients’ perceptions of COC arising from their experiences with mental health services. The qualitative inquiry confirmed several initial impressions we gained from the COC literature. First, most of the COC attributes identified in the literature reflected an administrative, service-provider conception of COC for this population (but see also 9,30). In short, the extant theoretical and research literature has ignored patient perceptions of COC. Second, the large number of attributes discarded through qualitative analysis confirmed our impression that COC is typically discussed as a normative construct. That is, COC is regarded as a worthy goal of mental health service administration, but specifics regarding its nature and how to develop continuity are less evident. The independent classifications of attributes derived from the literature showed a high degree of convergence. We regarded this as an indication that the various dimensions of COC (that is, domains and subdomains or themes) identified by the 2 investigator teams were valid. The consensus meeting of project coinvestigators brought a clinical perspective to the classification results, highlighting aspects of COC that were felt to be critical to service provision. The independent interviews with SPMI patients and their families provided convergent validation of the consensus meeting, because the same aspects of COC emerged as central consumer concerns. Notably, a 2001 consensus conference on COC (31) identified “core dimensions” of continuity across various patient groups that also showed good correspondence with the findings of our qualitative inquiry. The information, relational, care coordination, and longitudinality domains identified by conference delegates tended to be broader and more general than the aspects of COC outlined above, but the parallels are obvious and are consistent with our conceptualization of the construct. Comments from patients and their families also substantially confirmed the domains of COC attributes identified via the systematic literature review; that is, 92.5% of the attributes derived from interview data overlapped with those derived from the literature review. The interviews nonetheless provided a phenomenological perspective on COC that took into account the unique circumstances of individuals living with SPMI. Although not the focus of this paper, the interview transcripts consistently reflected the emotional intensity of patients’ experience with mental health service delivery. For example, many respondents expressed frustration, anger, and disenchantment with emergency hospital services and with the lack of linkages between hospital psychiatric services and community agencies. At the same time, most had positive comments about the efforts of health care workers and community support staff with whom they were involved. From the patients’ perspective, then, it appeared that the continuity afforded by their specific relationship to a provider or provider team was both critical and positive, whereas the continuity afforded by the mental health system as a whole was felt to be lacking. This differential weighting of provider and system contributions to COC, as perceived by patients, is an important area for future research. Our qualitative analysis confirmed the importance of both systemic interventions to promote COC and the specific patient–provider relationships underlying COC. Issues of accessibility and individualized care also emerged. However, the literature-derived attributes did not convey the importance of the affective dimension in patients’ experience of the system, reinforcing our argument that the literature on COC tends to be normative and oriented toward administrative issues, rather than descriptive and oriented to patient perceptions. The results of our qualitative inquiry also indicated that COC is a multidimensional construct, whether viewed in terms of a hierarchical model or in terms of thematic constituents. This contrasts with administrative client flow measures, which typically conceptualize COC according to a single dimension (for example, length of time between discharge and receipt of community services). A key question for our initial empirical use of the ACSS-MH is whether patient ratings will reflect multiple dimensions of the COC construct. In other words, will there be a correspondence between our conceptual formulation of COC and the structure of patient perceptions of continuity, as derived from ratings on the self-report questionnaire developed in this study? We suspect that there will be such correspondence, and if so, perceived continuity of care would represent a complex and multidimensional phenomenon. Beyond this basic question of correspondence, however, are issues associated with the value of a multidimensional measure of COC. A multidimensional measure is advantageous because it can lead to differential predictions about the impact of interventions designed to improve continuity. For example, we would predict that, if administrative interventions designed to improve COC were accurately perceived by patients, scores on items addressing problems of system integration would decrease, while scores on items addressing the patient–provider relationship, the responsiveness of treatment services, and the quality of provider actions would increase. In line with the hierarchical model presented in Figures 1 and 2, the draft ACSS-MH included items addressing both global and specific aspects of perceived COC in mental health services. Limitations and Future Directions Another possible limitation was our use of only a single sample of SPMI patients to derive potential scale items. The sample spanned the range of their involvement with services, encompassing both more acutely ill patients who had recently entered the mental health system and patients with chronic illness who had extensive involvement with treatment providers. It is unclear whether the reports by patients and their family members are generalizable to more homogeneous patient populations or to other locales. ConclusionsExisting concepts and measures of COC emphasize the objective characteristics of mental health service delivery. While important, such measures fail to take patient perspectives into account. To fill this gap in the literature, we used an inductive approach to develop a patient-centred measure of perceived COC. To our knowledge, the resulting ACSS-MH is the first measure of perceived COC addressing both the system and the provider to be presented in the mental health literature. We look forward to an examination of its psychometric properties and to further applications of the scale in mental health services research. NoteThe main literature search on continuity of care was conducted by Barbara Fraser, a graduate student in Public Health Sciences at the University of Alberta. The validation search of the literature was conducted by Judy Osborne, librarian for the Alberta Mental Health Board. We greatly appreciate their contributions to the work described in this article. Funding and SupportThis research was supported by a research program grant from the Canadian Health Services Research Foundation (grant number RC2-270-9) and by funding from the Alberta Mental Health Board, Eli Lilly Canada (unrestricted grant), the Alberta Heritage Foundation for Medical Research, and the Institute of Health Economics. References1. Saarento O, Oeiesvold T, Sytema S, Goestas G, Kastrup M, Loennerberg O, and others. The Nordic Comparative Study on Sectorized Psychiatry: continuity of care related to characteristics of the psychiatric services and the patients. Soc Psychiatry Psychiatr Epidemiol 1998;33:521–7. 2. Bachrach LL. Continuity of care and approaches to case management for long-term mentally ill patients. Hosp Community Psychiatry 1993;44:465–8. 3. Johnson S, Prosser D, Bindman J, Szmukler G. Continuity of care for the severely mentally ill: concepts and measures. Soc Psychiatry Psychiatr Epidemiol 1997;32:137–42. 4. Tessler RC, Willis G, Gubman GD. Defining and measuring continuity of care. Psychosocial Rehabilitation Journal 1986;10:27–38. 5. Eaton WW. Strategies of measurement and analysis. In: Knudson HK, Thornicroft G, editors. Mental health services evaluation. Cambridge (UK): Cambridge University Press; 1996. p 121–42. 6. Bachrach LL. Continuity of care for chronic mental patients: a conceptual analysis. Am J Psychiatry 1981;138:1449–56. 7. Kopelowicz A, Wallace CJ, Zarate R. Defining and measuring continuity of care. Psychosocial Rehabilitation Journal 1998;49:1313–6. 8. Sytema S, Micciolo R, Tansella M. Continuity of care for patients with schizophrenia and related disorders: a comparative South Verona and Groningen case-register study. Psychol Med 1997;27:1355–62. 9. Ware NC, Tugenberg T, Dickey B, McHorney CA. An ethnographic study of the meaning of continuity of care in mental health services. Psychiatr Serv 1999;50:395–400. 10. Adair CE, McDougall GM, Beckie A, Joyce A, Mitton C, Wild TC, and others. History and measurement of continuity of care in mental health services and evidence of its role in outcomes. Psychiatr Serv 2003;54:1351–6. 11. Bass RD, Windle C. A preliminary attempt to measure continuity of care in a community mental health clinic. Community Ment Health J 1972;9:53–62. 12. Rogers J, Curtis P. The concept and measurement of continuity of primary care. Am J Public Health 1980;70:122–7. 13. Pugh TF, MacMahon B. Measurement of discontinuity of psychiatric inpatient care. Public Health Rep 1967;82:533–8. 14. Barbato A, Terzian E, Saraceno B, Barquero FM, Tognon G. Patterns of aftercare for psychiatric patients discharged after short inpatient treatment. Soc Psychiatry Psychiatr Epidemiol 1992;27:46–52. 15. Shern DL, Wilson NZ, Coen AS, Patrick DC. Client outcomes II: longitudinal client data from the Colorado treatment outcome study. Milbank Q 1994;72:123–48. 16. Meisler N, Santos AB, Rowland MD, Smith S, Molloy M, Tyson S. Bridging the gap between inpatient and outpatient providers using organizational elements of assertive community treatment. Adm Policy Ment Health 1997;25:141–52. 17. Boyer CA, Olfson M, Kellermann SC, Hansell S, Walkup J, Rosenfield S, and others. Studying inpatient treatment practices in schizophrenia: an integrated methodology. Psychiatr Q 1995;66:293–320. 18. Bachrach LL. Case management revisited. Hosp Community Psychiatry 1992;43:209–10. 19. No authors listed. Service delivery. J Psychiatry Neurosci 1994;19(Suppl 1);22–8. 20. Roos G. Pile sorting: “kids like candy.” In: De Munck VC, Sobo EJ, editors. Using methods in the field: a practical introduction and casebook. Walnut Creek (CA): AltaMira; 1998. p 97–110. 21. Ryan GW, Bernard HW. Data management and analysis methods. In: Denzin NK, Lincoln YS, editors. Handbook of qualitative research. 2nd ed. Thousand Oaks (CA): Sage; 2000. p 769–802. 22. Glaser F, Straus E. The discovery of grounded theory. Chicago (IL): Aldine; 1967. 23. Smith TE, Hull JW, Hedayat-Harris A, Ryder G, Berger LJ. Development of a vertically integrated program of services for persons with schizophrenia. Psychiatr Serv 1999;50:931–5. 24. Greene RT. A comprehensive mental health care system for prison inmates: retrospective look at New York’s ten-year experience. Int J Law Psychiatry 1988;11:381–9. 25. ten Horn G, Giel R. Uses of a psychiatric case register: evaluation studies. Acta Psychatrica Belgica 1986;86:362–9. 26. Lindholm H. Sectorized psychiatry: a methodological study of the effects of reorganization on patients treated at a mental hospital. Acta Psychiatr Scand 1983;304:1–127. 27. McCartney PL. Psychiatric emergency services: issues and choices. Adm Policy Ment Health 1994;22:159–67. 28. Hoencamp E. Patient service requests. Psychiatr Serv 1996;47:1128–9. 29. Prendergast PJ. Psychiatric rehabilitation and case management in schizophrenia. In: Shriqui CL, Nasrallah HA, editors. Contemporary issues in the treatment of schizophrenia. Washington (DC): American Psychiatric Press; 1995. p 703–12. 30. Wallace T, Robertson E, Millar C, Frisch SR. Perceptions of care and services by the clients and families: a personal experience. J Adv Nurs 1999;29:1144–53. 31. Reid R, Haggerty J, McKendry R. Defusing the confusion: concepts and measures of continuity of healthcare. Final report. Prepared for the Canadian Health Services Research Foundation, the Canadian Institute for Health Information, and the Advisory Committee on Health Services of the Federal/Provincial/Territorial Deputy Ministers of Health. March 2002. Available: www.chsrf.ca/final_research/commissioner_research/programs/pdf/cr_contcare_2.pdf Author(s)Manuscript received February 2003, revised, and accepted October 2003. 1. Professor, Department of Psychiatry, University of Alberta, Edmonton, Alberta. 2. New Investigator, Canadian Institutes of Health Research; Investigator, Alberta Heritage Foundation for Medical Research (Population Health); Assistant Professor, Centre for Health Promotion Studies and Department of Public Health Sciences, University of Alberta, Edmonton, Alberta; Affiliate Scientist, Centre for Addiction and Mental Health, Toronto, Ontario. 3. Adjunct Associate Professor, Departments of Community Health Sciences and Psychiatry, University of Calgary, Calgary, Alberta. 4. Professor Emeritus, Department of Psychiatry, University of Calgary, Calgary, Alberta. 5. Senior Administrative Psychiatrist, Edmonton Mental Health Clinic, Edmonton, Alberta; Clinical Professor, Department of Psychiatry, University of Alberta, Edmonton, Alberta. 6. Clinical Professor, Department of Psychiatry, University of Alberta, Edmonton, Alberta. 7. Research Associate, Alberta Mental Health Board, Calgary, Alberta. 8. Research Associate, Alberta Mental Health Board, Edmonton, Alberta. Address for correspondence: Dr AS Joyce, Department of Psychiatry, University of Alberta, 8440 112 Street, Edmonton, AB e-mail: ajoyce@ualberta.ca
1 | 2
|
||||||||||||||||||||||||||||||||
