Needs and Appropriateness of Help According to Types of Professionals and Their Users



Marie-Josée Fleury, PhD
Assistant Professor, Department of Psychiatry, McGill University
Douglas Hospital Research Centre, Montreal, Canada

Guy Grenier, PhD
Research Associate, Douglas Hospital Research Centre, Montreal, Canada




Fleury M.J. & Grenier G. (2007). Needs and Appropriateness of Help According to Types of Professionals
 and Their Users.  International Journal of Psychosocial Rehabilitation. 12 (1), 5-22

Contact:
Marie-Josée Fleury, Ph.D.
Douglas Hospital Research Centre
6875 LaSalle Blvd.
Montreal, Quebec, Canada H4H 1R3
E-mail: flemar@douglas.mcgill.ca

Acknowledgements :
The Canadian Health Services Research Foundation (CHSRF), the Conseil québécois de la recherche sociale (CQRS), Fonds de recherche sur la société et la culture (FRSC), the Quebec department of health and social services, the Institut national de santé publique du Québec and the regional agencies of Quebec City, Bas-St-Laurent, Chaudière-Appalaches, Montréal-Centre and Outaouais all provided grant support for this study. We also express our thanks to Youcef Ouadahi for his help in collecting and analysing the data for the purposes of this article, and to Dr. Alain Lesage for his review of our article.

Abstract
The discrepancies in needs-identification and the appropriateness of help according to patients’ profiles, types of healthcare professionals, and users are examined based on 159 pairs of professionals and users. The Camberwell Assessment of Need (CAN) was used to assess: the number, severity and diversity of needs according to professionals and users, their degree of agreement and the appropriateness of help in meeting needs. The results show that patients tend to select as their primary healthcare provider a type of professional in accordance with their profile. Types of professionals tend to prioritize or neglect various needs in accordance with their background. The user/professional pairs with the highest agreement were those that obtained the most appropriate help. No type of professional can then identify all of its patients’ needs. More cross-vocational training programs as well as interdisciplinary needs-evaluation and diversified care, which include systematically the patient’s perspective, are thus necessary.

Keywords: needs-assessment; agreement between professionals and users; appropriateness of help; mental health types of professionals; patient profiles

Introduction
In the past several years, considerable consensus has emerged in the field of mental health on the necessity of considering the patient’s needs as a starting point for service planning. Evaluating patients’ needs has become a key component of the rehabilitation process (Andresen, Caputi, & Oades, 2000). Numerous evaluation tools have been devised to support this development; of these, the Camberwell Assessment of Need (CAN) is the most frequently used (Phelan, Slade, Thornicroft, Dunn, Holloway, Wykes et al., 1995).  Most studies using the CAN instrument have demonstrated moderate agreement between users and professionals in needs identification (Slade, Phelan, & Thornicroft, 1998; Slade, Leese, Ruggeri, Kuipers, Tansella, & Thornicroft, 2004). Other instruments have produced similar results (Comtois, Morin, Lesage, Lalonde, Likavëanova, & L’Écuyer, 1998; Gibbons, Bédard, & Mack, 2005). Authors explain the divergence in needs-perception by pointing to the different factors that affect users and professionals. In their view, socio-cultural environment, education, and past experiences play a primary role in users’ identification of their needs, whereas professionals are more influenced by training and professional values (Nielsen, Middleboe, Werdelin, Petersen, Mackeprang, Mortensen et al. 1999; Slade, 1996). Chaplain and Perkins (1999) state that the perceptions of people in the same profession are more closely aligned than with those of professionals from other disciplines. For instance, two nurses will tend to think alike more than, say, a nurse and a psychiatrist or a nurse and a psychologist. The same observation has been made of other health professionals (Meredith, Wells, & Camp, 1994; Shao, Williams, Lee, Badgette, Aaronson, & Cornell, 1997), even when they come from different countries (e.g. Australia and Singapore) (Parker, Chen, Jua, Loh, & Jorm, 2000).  However, no study has attempted to explain the divergence in the perceptions of types of professionals in relation to their patients’ clinical and socio-demographic profile and service-utilization. Correlations between needs-identification and the adequacy of services have also been neglected. This paper, drawn from a larger study on integrated service networks in Quebec, Canada (Fleury, Mercier,  Lesage, Ouadahi, Grenier, Aubé et al., 2004), aims to: (1) examine patients’ clinical and socio-demographic profile and service-utilization according to the profession or occupation of their most significant professional or mental healthcare provider; (2) assess the level of agreement in the number, severity and diversity of the needs identified by professionals in accordance with their occupation and their respective patients; and (3) evaluate the appropriateness of the help provided by services according to types of professionals and users.

Method
Data collection
Needs-analysis was performed with the French-language version of the Camberwell Assessment of Need (CAN) instrument (Bonsack & Lesage, 1998), whose reliability has been tested in several countries (Slade, Phelan, Thornicroft, & Parkman, 1996; Arvidsson, 2003).  An important characteristic of the CAN is its capacity to encompass patients’ needs and the types of care provided by relatives and established services. The CAN includes five categories of needs and 22 domains. The categories of needs are: (1) Basic (accommodation, food, daytime activities); (2) Health (physical health, psychotic symptoms, psychological distress, safety to self, safety to others, alcohol, drugs); (3) Functioning (self-care, looking after the home, childcare, basic education, money); (4) Social (company, intimate relationships, sexual expression); and (5) Services (information on disorder and treatment, transportation, telephone and benefits). There are five sections in the CAN. In the first, for each of the 22 domains, the patient indicates their perceived need severity on a three-point scale (no problem=0, moderate problem due to help given=1, serious problem whether or not help is received=2). When a moderate or serious problem is reported, the patient must complete the other sections of the questionnaire pertaining to the level of help received from relatives or services (none=0, low=1, moderate=2, high=3); the level of perceived need for help from services (none=0, low=1, moderate=2, high=3); and finally, adequacy of that help (right type of help provided –  1=yes or 0=no) and satisfaction (related to the amount of help provided – 1=yes or 0=no). In every section, patients can also answer “I don’t know”, which is considered as missing information. The CAN is based on a semi-directed interview which takes 30 to 45 minutes, and can be administered at the patients’ residence or elsewhere to their convenience.

Additional data were drawn from patients’ records, related to their socio-demographic (gender, age, income, education and current type of residence) and clinical profile (principal and secondary diagnosis, and life span: number of suicide attempts, criminal record, drug and alcohol abuse, and history of violence), and their use of health services (number and length of hospital stays between January 2002 to May 2004, number and type of resources used during the last year). In a brief complementary interview, patients and professionals were also asked to answer a question on their perception of the patients’ mental health state. Users also had to provide information on their service trajectory in the past year, such as the type and number of professionals involved in their care and the duration of their follow-up. The data collection has involved research assistants with a professional clinical background and trained for the purpose of the study.

Study design and cohort description
The study had a cross-sectional design based on patients living in six districts of the province of Quebec, Canada, chosen by convenience as most representative of semi-urban (areas 1, 2 and 3), rural (area 4) and urban (areas 5 and 6) settings. Population size ranged from 20,000 in area 4 to 128,000 in area 6 (Fleury, Grenier, & Lesage, 2006).  Targeted patients, aged 18 to 65 years, had to have been hospitalized in the last year and diagnosed with severe mental illness according to ICD-9 code 295 (Schizophrenia) or 297 (Delusional Disorder). Subjects were randomly selected from a list of 937 patients provided by hospital archives in the different districts. Sampling size was set at 30 patients per district, except for one rural district, set at 10, and an urban district at 60. From the list of eligible patients, we selected the first who agreed to participate in the study, for a global objective of 190. That number could not be reached in only one district. The final sampling was 186 patients, or 97.8% of the objective. It was representative of Quebec’ hospitalized population at the time of the data collection in terms of gender (χ2=3.97, df=1, p≤0.01), and age, with the exception of one urban territory (49.4; SD 10.3 versus 41.2; SD 12.6 for Quebec; χ2=29.52, df=4, p>0.05 – MED-ECHO, 2003-2004). Each participant signed a consent form approved by the relevant ethics boards.

The 186 patients selected were then asked to identify the professional whom they perceived as most significant and considered as their main mental healthcare provider; professionals identified in this manner were judged to be in the best position to answer the CAN. One hundred sixty-five professionals were identified. Professionals chosen by users included: 72 nurses (43.6%); 28 social workers (16.9%); 20 psychiatrists (12.1%); 22 educators (special or psycho-educators) (13.3%); 7 human relations officers (4.2%); 6 community workers (3.6%); 4 social agents (2.4%); 4 psychologists (2.4%); and 2 occupational therapists (1.2%). Nurses, social workers and community workers were present in the six districts covered by the study. Nurses were found mainly in urban district 6, where they constituted 64.4% of the significant professionals and in semi-urban district 1 (53.3%). Social workers were very few in number in districts 3, 5 and 6, and particularly present in the surrounding regions (1, 2 and 4). In district 2, they represented 40% of professionals chosen by users. Psychiatrists were more concentrated in district 5, where they made up 50% of the significant professionals. Most of the educators were found in district 3, while the community workers were found mostly in district 6. Finally, professionals’ profile in district 4 was heterogeneous, with no category of professional standing out. Community workers, human relations officers, and social agents have a university education in related social sciences branches (psychology, psycho-education, social work, communication, and criminology, etc.) and provide community follow-up. The difference in their appellations seems to be related principally to the organization with which they are affiliated (local community health centre, hospital or community resources). In this study, we refer to those professionals as “community workers.” Considering their small number, psychologists and occupational therapists have been excluded. Therefore, this study centres on 159 user/professional pairs.

A majority of the 159 patients were men (102 or 64.2%), whose mean age was 49.8 years (SD 11.8), single (109 or 68.6%), had a high-school education (85 or 53.5%), and were living autonomously in an apartment (86 or 54.1%). A diagnosis of schizophrenia was nearly the rule (155 or 97.5%). Nearly half of the cohort (79 or 49.7%) had a chronic syndrome such as mental retardation or personality disorder (as per DSM-IV). A history of family psychiatric problems was also reported for 72 patients (45.3%). Moreover, 49 patients (30.8%) had a history of problems with alcohol, 48 (29.1%) with drugs, 48 (30.2%) a history of violence, and 40 (25.2%) problems with the law. Fifty (31.4%) had tried to take their own life, averaging 1.80 lifetime suicide attempts (SD = 1.23). Users perceived their mental health state more positively (Figure 1) than professionals did (Figure 2). Only 16 users (10.1%) considered it to be bad, whereas 32 (20.1%) very good or excellent, compared to 39 professionals (24.5%) who estimated it to be bad and 13 (8.3%) very good. The average number of patient hospitalizations between January 2002 and May 2004 was 1.92 (SD = 1.47). The length of stay was less than 7 days for 53 users (33.3%) and between 1 and 2 weeks for 44 users (27.7%). On average, patients used 2.0 mental health resources in the past year (SD = 1.01), mainly outpatient clinics (111 or 69.8%), community resources (34 or 17.9%), local community health centres (28 or 17.6%), and specialized clinics (28 or 17.6%). Fifty-two (32.7%) mentioned being followed by a psychiatrist and a general practitioner. Twenty-nine (18.2%) had been seen by the same psychiatrist and 28 (17.6%) by the same general practitioner for more than ten years. Sixty (37.7%) reported consulting another professional as well, mainly a social worker (24 or 15.1%).





Statistical analyses of the data were conducted, using SPSS software (version 11.5). To measure the agreement between patients and professionals on overall needs-identification, the Kappa coefficient was used. Kappa measures the amount of agreement obtained between two raters beyond that which would be expected by chance alone (Spitznagel, & Helfzer, 1985; Crewson, 2005). Agreement is considered almost perfect when k ≥ 0.80; substantial if k is between 0.61 and 0.80; moderate, between 0.41 and 0.61; fair, between 0.21 and 0.40; and slight, between 0 and 0.20 (Middelboe, Mackeprang, Hansson, Werdelin, Karlsson,  Bjarnason et al., 2001).

Results
Users’ profile according to type of professionals (of the primary healthcare provider)
Patients who identified nurses as their most significant professionals had a socio-demographic profile similar to the patients’ general profile. Most of them were diagnosed with paranoid schizophrenia (34 out of 72 or 47.2%), and many had antecedents of violence (28 or 38.9%). The number of suicide attempts was rather high (28 or 38.9%). Of all the types of professionals, nurses’ perception of their patients’ mental health state was the most positive, with 78% rating it average to very good. Their users ranked second compared to patients of the other types of professionals, with 88.8% perceiving their mental health state from average to excellent. These patients were the main users of community follow-up. Twenty-nine out of 72 (40.2%) mentioned being followed by a psychiatrist and a general practitioner, 17 of them (23.6%) by the same general practitioner, and 16 (22.2%) by the same psychiatrist for more than ten years. Twenty-eight out of 72 (38.9%) reported consulting another professional as well, mainly a social worker (14 or 19.5%).

The social workers’ patients came mainly from rural or semi-urban settings (24 out of 28 or 85.7%). They were the only group with an almost equal number of women and men (12 patients out of 28 women or 42.9%). They also had the lowest rate of bachelors (15 out of 28 or 53.6%). They had mainly schizo-affective problems (13 out of 28 or 46.4%); a minority (4 out of 28 or 14.3%) also had another diagnosis. Psychiatric family antecedents (16 out of 28 or 57.1%) and suicide attempts (10 or 35.7%) were especially high. The social workers were also confronted with more cases of alcohol abuse than other professionals (13 out of 28 or 46.4%). Antecedents of violence were scarce (5 or 17.9%). The social workers’ perception of their patients’ mental health state was within the norm compared to overall professionals, with 68% considering it to be average to very good. Their patients’ perception was the second-highest of the user cohort, 92.8% evaluating it from average to excellent. These patients constituted the main users of resources (2.6 resources per user). Eighteen patients (64.3%) reported being followed by another professional, nine (32.1%) by a psychiatrist and a general practitioner – seven (25%) for more than ten years.

The psychiatrists’ patients were by far the most autonomous. Except for one, they lived in apartments, mostly in urban settings (13 or 65%), were rather young (averaging 42.6 years of age, SD 13.0), educated (12 out of 20 or 60% had completed junior college or university), and more than any other, single (16 out of 20 or 80%). They had fewer legal antecedents (3 or 15%), suicide attempts (also 3 or 15%) and chronic syndromes (7 or 35%). But, many had family psychiatric antecedents (15 or 75%), and a history of drug use (10 or 50%) and violence (7 or 35%). Psychiatrists perceived their patients’ mental state the most negatively, with 80% considering it average or bad. Conversely, 85% of their users judged it from average to very good; however, nobody in that group rated it excellent. The psychiatrists’ patients made scarce use of resources (1.8 per user), merely visiting outpatient clinics occasionally. Three users (15%) mentioned being followed by another professional as well as by a psychiatrist for the past ten years.

The educators’ patients differed from those of other professionals in many aspects. They were older (on average, 50.6, SD 7.7), a significant number were separated, divorced or widowed (7 out of 22 or 31.8%, against 33 out of 165 or 20% for the cohort). Many lived in intermediary resources (8 or 36.4%), and most in semi-urban settings (17 or 77.3%). The main diagnosis for these patients was schizo-affective disorder (13 or 59.1%), and a high number also had a chronic syndrome (17 or 77.3%). The rate of suicide attempts was the highest (8 or 36.4%). But, few patients had legal antecedents (4 or 18.2%), problems of violence (4 or 18.2%), and drug (3 or 13.6%) or alcohol abuse (2 or 9.1%). The educators’ perception of their patients’ mental health state was the second-lowest among all professionals, 77.2% judging it to be from average to bad. Conversely, 90.9% of their users evaluated it from average to excellent. The educators’ patients were the main users of specialized clinics. Nine (40.9%) reported being followed by another professional, mainly a social worker (6 or 27.3%). All were followed by a psychiatrist, and 7 (31.8%) by a psychiatrist and a general practitioner – two (9%) for more than ten years.

Finally, the patients who chose community workers were the youngest (averaging 39.3 years, SD 13.5), the most urban (13 out of 17 or 76.5%) and the least-educated (6 or 35.3% had completed elementary school only). These patients’ main diagnosis was paranoid schizophrenia (7 or 41.7%). They had few antecedents of violence (4 or 23.5%), legal problems (3 or 17.6%), or alcoholism (3 or 17.6%). Community workers ranked third in the perception of their patients’ mental health state, 64.7% evaluating it from average to very good. Community workers’ patients were the most numerous to judge it very good or excellent (47.0%), with only patient (5.9%) considering it bad. Community workers’ patients were among the most infrequent users of resources (1.8 per user), preferring services unrelated to mental health. Six patients (35.3%) were followed by another professional as well, generally a social worker (4 or 23.5%), and five (29.4%) by a psychiatrist and a general practitioner. One patient had been consulting the same general practitioner for more than ten years, and another, the same psychiatrist for as long.

Number, severity and diversity of needs according to professionals and users
Table 1 shows the number of problems – overall (total of moderate and serious) and serious – identified by professionals and users. Professionals reported that more than 40% of users had overall needs in the following areas: psychotic symptoms (126 or 79.2%); company (95 or 59.7%); daytime activities (84 or 52.8%); money (71 or 44.7%); psychological distress (69 or 43.4%); and looking after the home (66 or 41.5%). In needs-identification, self-care was the area where professionals and users most disagreed (44 versus 13 for users). More than 40% of users perceived problems in psychotic symptoms (124 or 78.0%), company (86 or 54.1%), food (77 or 48.4%), daytime activities (75 or 47.2%), money (72 or 45.3%), and psychological distress (71 or 44.7%). Information on disorder and treatment stood out as the area where users perceived the most problems as compared to professionals (62 versus 38 for professionals). As for serious problems, professionals and users agreed that they occurred more frequently in company, daytime activities, psychotic symptoms, intimate relationships, money, transportation, sexual expression, and psychological distress. In the 14 remaining areas, the proportion of users with serious problems was less than 10%. Professionals identified more serious problems in self-care (12 vs. 0) and basic education (10 vs. 4), whereas users identified them in information (15 vs. 3) and benefits (5 vs. 1).

Table 2 indicates the average number of needs (moderate, serious and overall) pointed out respectively by types of professionals and users, as well as the percentage of serious problems among overall problems. Needs identified by users averaged 5.97 (SD 3.40), of which 1.75 (SD 1.84) were serious (29.4%). As for professionals, the average number of overall problems they noted was 6.13 (SD 3.44) per patient, i.e. 4.18 moderate problems (SD 1.63) and 1.94 serious problems (SD 1.82). Serious needs thus represented nearly a third of all overall problems. Psychiatrists pointed out many fewer problems (moderate and serious) than any other type of professionals, with serious problems twice as low as the average. Conversely, community workers and social workers mentioned more moderate and serious problems than the average. Community workers, at a rate of 2.82 (SD 2.39), perceived more serious problems than any other type of professionals. Nurse showed converse profiles, mostly reporting moderate problems and few serious problems. Finally, educators listed needs nearest the average. Educators and community workers were the only ones to perceive more problems on average than their respective patients.

Table 2: Average number and percentage of needs according to types of professionals and their patients

Types of professionals/users	Moderate needs	Serious needs	Overall needs	Ratio needs Serious / Overall
Nurses     Users	4.50  4.64	1.87 2.14	6.37 6.78	29.4% 30.5%
Social workers   Users	4.54 5.46	2.25 1.93	6.79 7.39	33.1% 26.1%
Psychiatrists  Users	2.70 3.95	1.00 1.20	3.70 5.15	27.0% 23.3%
Educators  Users	4,14 2.95	2.00 1.36	6.14 4.31	32.6% 31.6%
Community workers Users	4.23 3.76	2.82 1.29	7.06 5.05	38.1% 25.6%
Professionals total	4.18	1.95	6.13	31.8%
Total needs per user	4.22	1.75	5.97	29.4%

Of the 22 CAN domains regarding overall needs: nurses noted fewer problems in company (40.0% in total, as compared with 45.3% of professionals); social workers noted fewer problems in transportation (just 10.0% of the problems; 17.6% in the view of professionals) and psychotic symptoms (14.7%). Only two domains caught the attention of psychiatrists: psychotic symptoms (identified by all of them) and company (13 out of 20 or 65.0%). Educators reported a higher number of problems in intimate relationships (25.0% in total; as compared with 13.8% for professionals), safety to self (21.1%), transportation (20.0%), benefits (16.7%), and safety to others (16.7%). Needs related to sexual expression (19.4% in total; 10.7% for professionals), safety to others (16.7%), basic education (16.7%), psychological distress (15.7%), drug use (15.0%), and intimate relationships (15.0%) were the main domains mentioned by the community workers. Two domains, childcare and telephone, were clearly mentioned less frequently by all types of professionals and reported mostly by the nurses.

As for serious needs, nurses were the only ones to identify at least one serious problem in each of the 22 CAN domains. They pointed out few serious problems in the social category of needs, but a great number in services and functioning. More than their patients, they mentioned serious needs in daytime activities (22 vs. 15) and transportation (13 vs. 9). But, they listed many fewer problems than their patients in intimate relationships (12 vs. 5), information on disorder and treatment (8 vs. 2), money (16 vs.11), sexual expression (12 vs. 8), and benefits (4 vs. 1). Psychiatrists reported problems in nine domains only. The few serious needs were found predominately in company, psychotic symptoms, and psychological distress. Conversely, social workers, educators, and community workers noted the highest percentage of serious problems. The most serious problems encountered by social workers, educators, and community workers were in the basic needs category. In comparison with their users, social workers pointed out many more serious problems in money (6 vs. 2), daytime activities (13 vs.10), and psychological distress (5 vs.2); and their patients, in company (13 vs. 9). Apart from the basic needs category, educators identified numerous problems in services and very few in functioning. They perceived more serious problems than their patients in daytime activities (13 vs. 5), company (11 vs. 5), intimate relationships (7 vs. 2), sexual expression (4 vs. 1), and psychotic symptoms (6 vs. 3). As for the community workers, they noted more serious problems than their patients in company (9 vs. 3), daytime activities (7 vs. 3), money (7 vs. 2), psychological distress (3 vs. 0), and sexual expression (4 vs. 0), whereas their users perceived more serious problems in information (2 vs. 0).

Agreement between types of professionals and their users on overall needs perceived
Table 3 shows the level of agreement between types of professionals and users based on the Kappa (k) coefficient. Globally, the degree of agreement between professionals and users was 0.42, a moderate rate. Nurses, social workers, psychiatrists and educators attained moderate agreement, with educators standing out, however, with a coefficient of 0.54. All types of professionals exhibited discordant to fair agreement in the needs domains of daytime activities, basic education, self-care, and telephone.

Table 3: Agreement between types of professionals and their users on overall needs perceived

CAN categories and domains of needs	Nurses (n=72)	Social workers (n=28)	Educators (n=22)	Psychiatrists (n=20)	Community workers (n=17)	Professionals total (n= 165)
Basic Accommodation	0.42 0.52	036 0.49	0.48 0.74	0.14 0.32	0.28 0.36	0.38 0.53
Food	0.47	0.38	0.50	-0.05	0.13	0.34
Daytime activities	0.22	0.23	0.22	0.35	0.23	0.24
Health      Physical health	0,54 0.42	0.45 0.29	0.70 0.79	0.61 0.63	0.28 0.46	0.52 0.49
Psychotic symptoms	0.29	0.01	0.46	0.00	-0.16	0.19
Psychological distress	0.50	-0.04	0.47	0.34	-0.16	0.31
Safety to self	0.18	0.63	0.70	-0.14	-0.09	0.30
Safety to others	0.16	-0.05	0.61	0.00	0.30	0.25
Alcohol	0.37	0.61	*	-0.05	0.64	0.45
Drugs	0.13	0.63	0.65	0.00	0.45	0.31
Social     Company	0.23 0.16	0.30 0.44	0.41 0.47	0,63 0.47	0.25 0.19	0.31 0.30
Intimate relationships	0.09	0.05	0.04	0.62	0.27	0.14
Sexual expression	0.23	0.18	0.46	0.00	0.00	0.21
Functioning  Looking after home	0.36 0.33	0.39 0.42	0.49 0.42	0.03 0.03	0.29 0.09	0.38 0.32
Self-care	0.23	-0.07	-0.08	0.00	0.21	0.14
Childcare	0.47	0.00	0.64	0.00	*	0.44
Basic education	0.11	0.00	*	*	0.35	0.16
Money	0.22	0.44	0.55	-0.09	0.03	0.30
Services   Information	0.25 0.06	0.41 0.13	0.39 0.32	0,56 0.77	0.11 -0.03	0.31 0.18
Transport	0.36	0.40	0.40	0.00	0.04	0.34
Telephone	0.31	0.00	*	*	0.00	0.24
Benefits	0.07	0.63	0.00	0.00	0.00	0.20
Total	0.41	0.42	0.54	0.47	0.27	0.42

* Kappa coefficient could not be calculated due to insufficiently spread data.

Nurses and their patients demonstrated moderate agreement in health and basic needs, but only fair agreement in the three other needs categories. They fared best in the domains of psychological distress, telephone and self-care – and worst in daytime activities. Social workers and their patients displayed fair agreement in all needs categories, except functioning and health. Specifically, self-care, psychotic symptoms, and psychological distress were needs domains where agreement was slight or discordant. But, it was substantial with respect to drugs, alcohol, safety to self and benefits. As for looking after the home, transportation and benefits, agreement was best achieved by the social workers as compared with the other types of professionals. Agreement between psychiatrists and users was substantial in social and health needs categories, fair in services, and slight in basic and functioning needs. In social needs, the domain of intimate relationships stood out when compared with other types of professionals. Psychiatrists also exhibited the highest level of agreement in the domains of company, daytime activities, and information on disorder and treatment. However, they exhibited slight or discordant agreement in five out of the seven health-needs domains and with regard to food (basic needs category). Educators were the only professionals who marginally presented a moderate rate in four needs categories. The level of agreement was high in the domains of accommodation, food, physical health, psychotic symptoms, safety to others, safety to self, drugs, company, sexual expression, looking after the home, childcare, money and transport, but fair in daytime activities, slight in intimate relationships, and discordant in self-care. The agreement between community workers and users was the lowest one, with a coefficient of 0.27, fair in all categories except in service needs where it was slight. But, it was substantial regarding alcohol use, and highest in identifying basic education problems.

Appropriateness of help in meeting needs
Professionals reported a rate of appropriate help of 84.8% (827 out of 975) vs. 74.7% (710 out of 950) for users. For professionals, the percentage of needs having received adequate help varies from 71.6% (53 out of 74) for psychiatrists to 91.7% (110 out of 120) for community workers. As for users, the percentage oscillates from 52.3% (45 out of 86) for community workers’ patients to 87.4% (83 out of 95) for educators’. The closest rates between users and professionals occur among users/educators (87.4% vs. 91.1% or 123 out of 135) and users/psychiatrists (70 out of 103 or 68.0% vs. 71.6%). The respective total of appropriate help for users/nurses (356 out of 482 or 73.9% vs. 384 out of 459 or 83.7%) and users/social workers (157 out of 207 or 75.9% vs. 169 out of 190 or 84.2%) is near the overall mean. Lastly, there is an important gap on help adequacy between community workers (91.7%) and their patients (52.3%).

Discussion
The aim of this article was to examine and explain the discrepancies in needs-identification and appropriateness of help according to patients’ profiles and types of healthcare professionals and their users. For that purpose, we interviewed 159 user/professionals pairs, using the French-language version of the Camberwell Assessment of Need (CAN) instrument.

Selection of a type of professional as the primary healthcare provider may be explained by users’ socio-demographic and clinical profile and resources utilization. Thus, the most autonomous patients, having fewer needs and using scarcer resources, were inclined to opt for a psychiatrist whereas patients living in intermediary resources, being older, presenting chronic syndromes and consulting more specialized clinics tended to choose an educator. Users grappling with numerous problems and utilizing more diversified resources selected nurses and social workers to a greater extent. Nurses were further correlated with patients diagnosed with paranoid schizophrenia and presenting antecedents of violence. Social workers were more favoured by users with schizo-affective disorders and alcohol abuse. Finally, patients preferring a community worker as their most significant professional were young, poorly educated, and lived for the most part in urban settings. These patients used fewer resources and usually considered their mental health state to be better than the rest of the cohort.

    The perception of users’ needs expressed by different types of professionals varied widely. Psychiatrists had a tendency to identify fewer moderate and serious problems than other professionals. Half of the needs occurred in the health category: in particular, psychotic symptoms. Company was the other domain to which psychiatrists paid more specific attention. Psychiatrists also viewed their patients’ state of mental health much more negatively than did other types of professionals. They seemed to associate mental health state more with clinical problems directly related to mental disease such as psychotic symptoms and psychological distress and to interpersonal issues (i.e. company). Conversely, social workers as well as community workers (for the most part made up of professionals with a social science background) listed more moderate and serious needs than other types of professionals. They also identified a greater number of overall problems in the social and functioning categories than in health and services needs. But, these two groups found fewer overall problems in psychotic symptoms, mentioning instead more needs in substance abuse for the first group, and safety to self and others for the second group. They broadly defined their patients’ mental health state, relating it more to basic and functioning needs categories, associated with the indirect effects of schizophrenia. Educators and to a lesser extent nurses outlined overall needs more evenly in the five categories. Nurses largely reported moderate problems in all categories, except social needs, but very few serious problems globally. Educators noticed numerous serious problems in all needs categories, save functioning. These results confirm Slade’s (Slade, 1994) hypothesis that healthcare providers tend to emphasize needs in certain domains in accordance with their professional background.

With such differences in needs-identification and in light of the Slade study mentioned above, agreement between professionals and users was expected to be difficult to reach. Moderate agreement between users and professionals (0.42) was obtained in our study. This corresponded almost exactly to what Slade et al. (Slade et al. 1998) had reported (0.44). An almost identical rate of moderate agreement was found among psychiatrists, social workers, and nurses. Only educators stood slightly out with an agreement of 0.54, while community workers exhibited a fair one (0.27). Agreement varied considerably from 0.31 in the social needs category to 0.52 in health needs, from 0.14 in the domains of intimate relationships and self-care to 0.51 in accommodation. The higher agreement between educators and users may be explained by the time educators spend daily with their patients in rehabilitation and reintegration. This allows them to be better aware of their patients’ current and evolving needs. The only other types of professionals as much in contact with their users are those making regular house calls, such as nurses or social workers in intensive community follow-up. According to  Macpherson, Varah, Summerfield, Foy, & Slade, (2003), the agreement between users and healthcare providers is higher when the professional is a clinician engaged in patient aftercare, therefore more aware of users’ needs. The low agreement between community workers and their patients may be explained by the users’ profile. These patients tend to misjudge or ignore their disease, as shown by the highly positive perception of their mental health state and the underestimation of their psychotic symptoms and psychological distress. Their mistrust of psychiatrists and other mental health specialists may testify to their paranoid schizophrenia. Being younger and less educated, they may also be at an earlier stage of recognition and acceptance of their mental disorder.

Regardless of profession, healthcare providers tended strongly not to recognize some of their patients’ needs, such as information on disorder and treatment and benefits. They also had difficulties assessing intimate relationships and sexual expression, as demonstrated by the slight agreement rate between their responses and their patients’. Middelboe, Mackeprang, Hansson, Werdelin, Karlsson, Bjarnason et al., (2001) noted that these four domains are overlooked by professionals because they are not directly related to schizophrenia, which explains the inappropriateness or insufficiency of care to answer those needs. Moreover, schizophrenia is associated with cognitive deficiencies that reduce capacity to store information (Brenner, Hodel, Roder, & Corrigan, 1992; Bengtsson-Tops, & Hansson, 1999), hence the necessity for a professional to be alert when answering patients’ questions regarding their mental health.
Psychotic symptoms presented slight agreement as well, but it was primarily due to substantial skews in the ratings’ distribution (Issakidis, & Teeson, 1999; Slade et al., 1996). Kappa statistics are unstable when a zero or a small number appears in some cells (Feinstein, & Cicchetti, 1990).  In fact, the agreement for psychotic symptoms was 71.7% for all professional/user pairs, ranging from 90% (18 out of 20) for psychiatrists to 68.7% for educators (15 out of 22).

Self-care and drugs as well as company and daytime activities and safety to self and others are also domains where important discrepancies were observed between users and professionals. Olfactory deficits often found among individuals with schizophrenia (Moberg, Doty, Turetsky, Arnold, Mahr, Gur et al. 1997; Hudry, Saoud, d’Amato, Daléry, & Royet, 2002) may account for the small number of patients mentioning self-care problems. Moreover, refusal to wash is one of the negative symptoms generally associated with schizophrenia. Drug users have a tendency to deny their substance abuse problems whether or not they suffer from schizophrenia. The low rate of agreement for company and daytime activities may be due to some users’ refusal to join structured or groups activities, especially involving patients only; however, it may also be due to the types of structured activities being inappropriate for patients’ needs. Finally, some patients underestimate the danger that they represent to themselves and to others.

In spite of the global moderate agreement between professionals and users, the patients’ needs were met at 74.7% according to users and at 83.8% according to professionals. Does this mean that agreement between users and professionals is not as important as is generally believed? The greater the similarity in user/professionals responses, the higher, generally, the percentage of appropriate help given. The sole exception is found among user/psychiatrist pairs, where, despite moderate agreement, the appropriateness of help is under average. Users and psychiatrists nevertheless rate the help adequacy similarly (68% vs. 71.6%). The main dissatisfactions of the psychiatrists’ patients pertain to the length of the consultations and intervals between appointments. As for community workers, their patients’ profile seems to account for the poor results achieved. Their patients’ unmet needs are generally associated with poverty (food, money, basic education, etc.). Moreover, users who identified a community worker as their most significant professional are less inclined to ask for help. They ask for moderate or high level of help for only 30 of their 86 problems (34.9% vs. 46.9% or 446 out of 950 for all users). They also receive less moderate or important (high) help from local services (25.6% or 22 out of 86 vs. 42.0% or 399 out of 950) and from their relatives (13.9% or 12 of 86 vs. 19.7% or 188 out of 950) than the general mean. It would be unfair to associate the gap between community workers and their patients with a lack of competence in this category of professionals. In fact, community workers provide some help to a marginal population which, without them, would be deprived of mental health services.

To sum up, our results show the importance of patients’ profiles and types of professionals in needs-assessment. Patients tend to select as their primary healthcare provider a type of professional more closely related to their socio-demographic, clinical and resources utilization profile with regard to the number, severity and diversity of their needs. Types of professionals are more inclined to prioritize or neglect certain needs depending on their own background. The moderate agreement in overall needs and the considerable appropriateness of help between professionals and users obtained in our study seems testify this trend.

Conclusion
There are certain shortcomings in this study, namely the lack of complete data on the most significant professionals’ years of practice and users’ length of care with them. These two elements play an important role in understanding patients’ needs. Moreover, the size of our sampling, specifically for the community workers as well as the lack of experience of some of the CAN interviewers may have slightly biased certain parts of the data. Also, some variables that may have played a role in the appropriateness of help have not been considered here, including: help given respectively by relatives or services; level of help needed; extent of continuity of care; and network services integration. Moreover, it would be interesting to analyse the assessments of different professional disciplines on the same patients, or to observe the degree of agreement between healthcare providers and users at two measuring times.

Nevertheless, to our knowledge, this study is the first to provide an exhaustive comparison between the perceptions of different types of professionals of their patients’ needs and to relate them to their patients’ profiles. The results of this study reinforce the importance of considering the viewpoints of both professionals and users in needs-assessment for a more comprehensive perspective. In spite of all the efforts made these past few years to promote the bio-psychosocial model in the mental health field, this study shows that professionals still prioritize some needs to the detriment of others. No professional may claim to know perfectly his or her patient’s condition. It is therefore important that mental health services offer more cross-vocational training programs as well as interdisciplinary needs-evaluation and diversified care for patients – systematically including patients’ perspective.

Andresen, R., Caputi, P., & Oades, L.G. (2000).  Interrater reliability of the Camberwell Assessment of Needs Short Appraisal Schedule.  Aust NZ J Psychiatry 34 (5): 856-861.

Arvidsson, H. (2003). Test-retest reliability of the Camberwell Assessment of Need (CAN).  Nord J Psychiatry 57: 279-283.

Bengtsson-Tops, A., & Hansson, L. (1999).  Clinical and social needs of schizophrenic outpatients living in the community: the relationship between needs and subjective quality of life.  Social Psychiatry Psychiatr Epidemiol 34: 513-518.

Bonsack, C., & Lesage, A.  (1998).  Deux instruments pour évaluer les besoins de soins de santé mentale : une étude comparative chez des personnes longuement hospitalisées.  Ann Med Psychol (Paris) 156 (4): 244-257.

Brenner, H.D., Hodel, B., Roder, V.,& Corrigan, P. (1992).  Treatment of cognitive dysfunctions and behavioural deficits in schizophrenia.  Schizophr Bull 18: 21-26.

Chaplin, R., & Perkins, R. (1999).  HoNOS: A cautionary tale of their use in a rehabilitation service.  Psychiatric Bulletin 23: 20-21. 

Comtois, G., Morin, C., Lesage, A., Lalonde, P., Likavëanova, E., & L’Écuyer, G. (1998).  Patients versus rehabilitation practionners: a comparison of assessments of needs for care.  Can J Psychiatry 43: 159-165.

Crane-Ross, D., Roth, D., & Lauber, B.G. (2000).  Consumers’ and case managers’ perceptions of mental health and community support service needs.  Community Ment Health J 36: 161-178.

Crewson, P.E. (2005).  Reader Agreement Studies.  AJR 184: 1391-1397.

Feinstein, R., & Cicchetti, D.V.  (1990).  High agreement but low kappa: I. The problems of two paradoxes.  J C Epidemiol 43: 543-549.

Fleury, M.J., Grenier, G., & Lesage, A. (2006).  Agreement between staff and service users concerning the patientele’s mental health needs: a Quebec study. Can J Psychiatry 51: 281-286.

Fleury, M.J., Mercier, C., Lesage, A., Ouadahi, Y., Grenier, G., Aubé, D. et al. (2004).  Réseaux intégrés de services et réponse aux besoins des personnes avec des troubles graves de santé mentale. Fondation Canadienne de recherche sur les services de santé, Ottawa.

Gannon, M., Ryan, M., & Kinsella, A. (2000).  Assessment of need in patients according an inner city psychiatric service. Ir J Psychol Med 17: 25-28.

Gibbons, C., Bédard, M., & Mack, G. (2005).  A comparison of patient and mental health worker assessment of needs and unmet needs.  J Behav Health Serv Res 32: 95-104.

Hansson, L., Vinding, H.R., Mackeprang, T., Sourander, A., Werdelin, G., Bengtsson-Tops, A., Bjarnason, O., Dybbro, J., Nilsson, L., Sandlund, M., Sörgaard, K., & Middelboe, T. (2001).  Comparison of key worker and patient assessment of needs in schizophrenic patients living in the community: a Nordic multicentre study.  Acta Psychiatr Scand 103: 45-51.

Hudry, J., Saoud, M., d’Amato, T., Daléry, J., & Royet, J.,P. (2002).  Rating of different olfactory judgements in schizophrenia. Chem.  Senses 27: 407-416.

Issakidis, C., & Teeson, M. (1999).  Measurement of need for care: a trial of the Camberwell Assessment of Needs and the Health of the Nation Outcome Scales.  Aust NZ J Psychiatry 33: 754-59.

Lasalvia, A., Ruggeri, M., Mazzi, M.A., & Dall’Agnola, R.B. (2000).  The perception of needs for care in staff and patients in community-based mental health services. The South-Verona Outcome Project 3.  Acta Psychiatr Scand 102: 366-75.

Macpherson, R., Varah, M., Summerfield, L., Foy, C., & Slade, M. (2003).  Staff and patient assessments of need in an epidemiologically representative sample of patients with psychosis.  Social Psychiatry Psychiatr Epidemiol 38: 662-67.

Meredith, L.S., Wells, K.B., & Camp, B. (1994).  Clinician speciality and treatment style for depressed outpatients with and without medical comorbidities. Arch Fam Med 8: 116-121.

Middelboe, T., Mackeprang, T., Hansson, L., Werdelin, G., Karlsson, H., Bjarnason, O. et al. (2001).  The Nordic Study on schizophrenic patients living in the community. Subjective needs and perceived help.  European Psychiatry 16: 207-15.

Moberg, P.J., Doty, R.L., Turetsky, B.I., Arnold, S.E., Mahr, R.N., Gur et al. (1997).  Olfactory identification deficits in schizophrenia: correlation with duration of illness.  Am J Psychiatry 154: 1016-1018.

Nielsen, L.F., Middleboe, T., Werdelin, G., Petersen, L., Mackeprang, T., Mortensen, E.L. et al. (1999).  Inter-rater reliability and test-retest reliability of the Danish version of the Camberwell Assessment of Needs.  Nord J Psychiatry 53: 371-376.

Parker, G., Chen, H., Jua, J., Loh, J., & Jorm, A.F. (2000).  A comparative mental health literacy survey of psychiatrists and other mental health professionals in Singapore. Aust NZ J Psychiatry 34: 627-636.

Phelan, M., Slade, M., Thornicroft, G., Dunn, G., Holloway, F., Wykes, T. et al.  (1995).  The Camberwell Assessment of Needs: the validity and reliability of an instrument to assess the needs of people with severe mental illness.  Br J Psychiatry 167: 589-595.

Shao, W.A., Williams, J.W., Lee, S., Badgette, R.G., Aaronson, B., & Cornell, J.E. (1997).  Knowledge and attitudes about depression among non-generalists and generalists. J Fam Pract 44: 161-168.

Slade, M. (1994).  Needs assessment involvement of staff and users will help to meet needs. Br J Psychiatry 165: 293-296.

Slade, M., Leese, M., Ruggeri, M., Kuipers, E., Tansella, M. & Thornicroft, G. (2004) Does meeting need improve quality of life?  Psychother Psychosom 73: 183-189.

Slade, M., Phelan, M., Thornicroft, G., & Parkman, S. (1996).  The Camberwell Assessment of Need (CAN): comparison of assessments by staff and patients of the needs of the severely mentally ill. Soc Psychiatry Psychiatr Epidemiol  31: 109-13.   

Slade, M., Phelan, M., & Thornicroft, G. (1998).  A comparison of needs assessed by staff and by an epidemiologically representative sample of patients with psychosis. Psychol Med 28: 543-50.

Spitznagel, E.L., & Helfzer, J.E.  (1985).  A proposed solution to the base rate problem in the kappa statistic. Arch Gen Psychiatry 42: 725-728.   

Wiersma, D., Nienhuis, F.J., Giel, R., & Slooff, C.J. (1998).  Stability and change in needs of patients with schizophrenic disorders: a 15- and 17- year follow-up from fist onset of psychosis, and a comparison between ‘objective’ and ‘subjective’ assessments of needs for care. Soc Psychiatry Psychiatr Epidemiol 33: 49-56.