met and unmet needs of schizophrenia patients in a spanish sample

Met and Unmet Needs of SchizophreniaPatients in a Spanish Sample

by S. Ochoa, J.M. Haro, J. Autonell, A. Pendas, F. Teba, M. Marque?:,and the NEDES Qroup

AbstractDeinstitutionalization of people with schizophreniaincreases the importance of evaluating their needs.This study set out to identify the most common needsof people with schizophrenia who live in the commu-nity, analyze how those needs differ when evaluated bystaff or by patients, describe the kind of help patientsreceive, and find out the variables that correlate withhaving unmet needs. A random sample of 231 outpa-tients with schizophrenia were evaluated with theCamberwell Assessment of Need and other predictorand outcome variables. Staff detected more needs thanpatients did. Mean number of needs as rated bypatients was 5.36 and staff 6.6 (p < 0.001). Mean num-ber of unmet needs was also greater when assessed bystaff than by patients: 1.38 versus 1.82 (p < 0.001). Themost frequently detected needs by patients involvedpsychotic symptoms, house upkeep, food, and infor-mation. Staff most often detected needs involving psy-chotic symptoms, company, daytime activities, houseupkeep, food, and information. In a multiple regres-sion model, needs were weakly associated with theclinical variables and quality of life. Needs assessmentis complementary to clinical evaluation in schizophre-nia.

Keywords: Schizophrenia, needs assessment, psy-chopathology, disability, quality of life.

Schizophrenia Bulletin, 29(2):201-210,2003.

The development of a system of comprehensive commu-nity services and the deinstitutionalization process startedin Barcelona at the beginning of the 1980s. The systemthat was created is organized into sectors of around100,000 inhabitants. Each sector has a mental health carecenter (MHCC) that is the central element of the commu-

nity's mental health care. The MHCC has an intense rela-tionship with primary health care and other psychiatricservices. Besides the MHCC, each sector either has or hasaccess to a number of psychiatric services, including daycenters, day hospitals, acute inpatient units, and long-stayinpatient units. Sheltered accommodation and shelteredwork are still uncommon in Barcelona. Community ser-vices are either public or publicly funded, and all citizenshave access to care. Although services have been muchimproved during the past 2 decades, many gaps still exist.

The main duty of public mental health care services inSpain is caring for the most severely ill, many of thempeople with schizophrenia. When treatment plans for peo-ple with schizophrenia are designed, two complementaryapproaches can be taken. First, designers can focus on thepatient's disabilities or handicaps and identify which ser-vices or elements are most appropriate to help the patientovercome those limitations. The second, more comprehen-sive alternative, one more focused toward specific inter-ventions, involves identifying the patient's needs. Needshave been defined as "the requirements of individuals toenable them to achieve, maintain or restore an acceptablelevel of social independence or quality of life" (Depart-ment of Health Social Services Inspectorate 1991). In psy-chiatry, need is employed to inform service provision andplan individual care (Slade 1994). Therefore, a more use-ful definition of need would be "the ability to benefit insome way from health care" (Stevens and Gabbay 1991).Need is a dynamic and context-dependent concept (Nettenand Beecham 1993; Slade et al. 1996; Andrew and Hen-derson 2000).

Send reprint requests to Dr. S. Ochoa, Unitat de Formaci6 i Investigari6,Sant Joan de Dgu-Serveis Salut Mental, C/ Dr. Pujades, 42, Sant Boi deLlobregat, Barcelona, Spain; e-mail: [email protected].

201

by guest on July 2, 2016http://schizophreniabulletin.oxfordjournals.org/

Dow

nloaded from

http://schizophreniabulletin.oxfordjournals.org/

Schizophrenia Bulletin, Vol. 29, No. 2, 2003 S. Ochoaetal.

Before deinstitutionalization, patients with severeschizophrenia lived in hospitals and most of their basicneeds were met by the institutions. Nowadays in Spain,with admission to long-stay inpatient units very uncom-mon, patients live in the community, most of them withtheir parents or siblings (Haro et al. 1998). Community ser-vices, family members, and friends take care of patientneeds. But it is not clear how and to what degree they do it.

Although several instruments have been developed toevaluate needs (Needs for Care Assessment Schedule[Brewin and Wing 1987]; Cardinal Needs Schedule [Mar-shall 1994]), during recent years the most widely used hasbeen the Camberwell Assessment of Need (CAN) (Phelanet al. 1995). The CAN allows us to determine patient andstaff perception of needs in 22 areas. It was designed toinform researchers and mental health professionals in theassessment of patient service needs.

Different studies published in Europe have identifiedvarious numbers and types of needs. Most studies that usethe CAN have found that patients detect more needs thanstaff (Slade et al. 1996, 1998, 1999). In a sample of psy-chotic outpatients, Slade et al. (1998) found a mean numberof needs of 6.1 per patient when assessed by staff, and 6.7when assessed by patients. The areas of need more fre-quently reported by staff were psychotic symptoms, day-time activities, company, and transport; by patients, theywere psychotic symptoms, transport, daytime activities,company, and food. In a study from the Netherlands(Wiersma et al. 1998), the most common needs identifiedby patients were psychological distress, intimate relation-ships, sexual expression, and daytime activities. Using asample of schizophrenia outpatients from Nordic countries,Hansson et al. (2001) reported that staff detected moreneeds than patients did. Staff-assessed needs mainlyinvolved company, psychotic symptoms, and daytimeactivities, and patient-assessed needs involved company,intimate relationships, and psychological distress.McCrone et al. (2001) found varying numbers of needsamong the cities studied, with Amsterdam having the high-est level of need and Santander (Spain) the lowest. Some ofthese differences in type and number of needs could beexplained by differences in sample selection, because therewas no homogeneity in diagnosis and treatment context(Thornicroft et al. 1996). Agreement in need detection bystaff and patients is said to be better in areas where there isa specific service for a concrete need (Slade et al. 1998).

The objectives of our study were to identify the mostcommon needs of schizophrenia patients who live in thecommunity, analyze how the needs differ when evaluatedby staff or by patients, describe what kind of help andwhose help people with schizophrenia receive for coveringtheir needs, and find the sociodemographic, social, andclinical correlates of the presence of needs.

Patients and Methods

Two hundred and thirty-one persons with schizophreniawere randomly selected from a computerized register thatincluded all patients under treatment in the five mental healthcare centers that participated in the study. The registerincluded personal information (name, address, identificationnumber, etc.), sociodemographic information (date of birth,gender, etc.), diagnosis, and the data from and type of visitsfor all patients that had ever been seen at the centers. Thefive catchment areas (Cerdanyola, Ciutat Vella, Cornelia,Gava, El Prat) contain a population of 440,000 adults fromthe city of Barcelona and its surroundings and house peopleof different sociodemographic backgrounds.

Inclusion criteria were (1) to have a primary diagnosisof schizophrenia according to DSM-FV criteria, (2) to bebetween 18 and 65 years old, (3) to live in the catchmentarea, and (4) to have had at least an outpatient visit duringthe 6 months prior to the beginning of the study. Patientswith a diagnosis of mental retardation or neurological dis-order were excluded.

The diagnosis of all the selected individuals wasreviewed by their treating psychiatrist. In case of discrep-ancy between the treating psychiatrist's diagnosis and thediagnosis of schizophrenia that was in the register, the casewas evaluated by two psychiatrists to make a final deci-sion regarding the inclusion of the patient.

Selected individuals were informed by their psychia-trist about the objectives and methodology of the studyand provided verbal informed consent to participate.

Evaluation

All patients were evaluated with the following instru-ments:

1. A sociodemographic and clinical questionnaire thatincluded information on psychiatric history and comor-bidity

2. The Positive and Negative Syndrome Scale (PANSS),Spanish version (Kay et al. 1986; Peralta and Cuesta1994)

3. The Global Assessment of Functioning Scale (GAP),Spanish version (Endicott et al. 1976; APA 1995)

4. The Disability Assessment Schedule (DAS), short ver-sion (Sartorius et al. 1986; World Health Organization1992)

5. The Quality of Life questionnaire (QOL), Spanish ver-sion (Baker and Intagliata 1982; Bobes et al. 1995)

6. The CAN, Spanish version (Phelan et al. 1995; Rosales1999; McCrone 2000). The CAN evaluates the presenceof needs in 22 areas: accommodation, food, houseupkeep, self-care, daytime activities, physical health,

202


Dow

nloaded from


Patients in a Spanish Sample Schizophrenia Bulletin, Vol. 29, No. 2, 2003

psychotic symptoms, information, psychological dis-tress, risk to self, risk to others, alcohol, drugs, com-pany, intimate relationship, sexual expression, childcare, education, telephone, transport, money, and bene-fits. In each of these areas the CAN determines whethera need exists, whether it is met, who provides the help(formal and informal care), and whether the help isappropriate. The questionnaire is actually two in onebecause it is evaluated by both the staff members whotreat the patients and the patients, independently.

The sociodemographic and clinical questionnaire, thePANSS, the DAS, and the GAP were administered by thepatient's treating psychiatrist. The CAN and the QOL werecompleted by the center's social worker. All the evaluatorsparticipated in a training course.

Statistical Analysis

We first report on descriptive statistics of met and unmetneeds. Cohen's kappa was calculated to determine theagreement between patients and staff when rating the pres-ence of needs. According to Landis and Koch (1977), akappa coefficient of 0.4 to 0.6 indicates moderate agree-ment, 0.6 to 0.8 indicates substantial agreement, and 0.8 to1.0 indicates almost perfect agreement. Comparisonbetween number of needs assessed by patients and staffwas made with the Wilcoxon matched pairs test.

To determine the association of total number of needsand number of unmet needs with the sociodemographicand clinical variables, a stepwise multiple linear regres-sion was used. Sex, age (categorized in three strata), mari-tal status, age of onset, the PANSS, the GAF, the DAS, andthe QOL were included in the models.

To study the influence of sociodemographic and clini-cal factors on the presence of each need, a logistic regres-sion analysis was performed. Twenty-two models, one perneed as dependent variable, were fitted. Independent pre-dictor variables were gender, marital status, age, age ofonset, length of illness, PANSS subscales, the GAF, theDAS, and the QOL. Before regression, exploratory analy-sis was performed. Variables that showed a relationshipwith need were entered into the model one by one. Onlythose variables showing statistical significance using max-imum likelihood estimates were fitted. A final model isreported. All statistical analyses were calculated withSPSS for Windows 6.0 (Norusis 1993).

Results

Table 1 shows the characteristics of the patients includedin the study. Approximately two-thirds of them were male.

Table 1. Demographic and clinical characteristicsof the sample (n = 231)

Characteristic

Male, n (%)

Marital status, n (%)

Single

Married

Divorced

Widowed

Living situation, n (%)

Alone

With parents

With other family members

With spouse/children

In residence/sheltered housing

In other situation

Occupational status, n (%)

Employed

On permanent sick leave

With other occupational status

Age, mean (SD)

Age at onset, mean (SD)

147 (63.6)

154(66.7)

49 (21.2)

26(11.3)

2 (0.9)

29(12.6)

117(50.6)

15(6.5)

58(25.1)

9 (3.9)

3(1.3)

19(8.2)

153(66.2)

59 (25.6)

39(12.1)

23 (7.4)

Note.—SD = standard deviation.

Patients were more often single, lived with parents, andwere not working because they were on permanent sickleave. Mean age was 39 years, and mean length of illness15 years. The evaluation was completed for 195 patients,84.4 percent of the total sample. There were no significantdifferences in any of the sociodemographic variablesshown in table 1 between the people who answered thequestionnaire and those who did not.

Staff detected more needs than patients did. Meannumber of needs as rated by patients was 5.36 (standarddeviation [SD] = 2.71), while staff detected a mean of 6.6(SD = 3.17) per patient. These differences are statisticallysignificant (p < 0.001). Mean number of unmet needs wasalso greater when assessed by staff than by patients: 1.38(SD = 1.75) versus 1.82 (SD = 1.98) (p < 0.001).

The most frequently detected needs by patientsinvolved psychotic symptoms, house upkeep, food, andinformation (table 2). Staff most often detected needs inthe areas of psychotic symptoms, company, daytime activ-ities, house upkeep, food, and information (table 2). Themain discrepancies arose in company, daytime activities,psychotic symptoms, and self-care. Physical health, psy-chological distress, and benefits were the only areas inwhich patients detected more needs than staff. Unmet need

203


Dow

nloaded from


Table 2. Percentage of patients with need and unmet need as rated by staff and patients, agreement between their ratings, andpercentage of needs detected by only staff or patient

g

Accommodation

Pood

House upkeep

Self-care

Daytime activities

Physical health

Psychotic symptoms

Information

Psychological distress

Risk to self

Risk to others

Alcohol

Drugs

Companionship

Intimate relationship

Sexual expression

Child care

Education

Telephone

Transport

Money

Benefits

Presenceof needrated bystaff (%)

9.2

51.8

57.4

27.7

54.9

23.6

96.4

46.7

31.8

13.3

6.2

16.4

9.7

58.5

26.2

14.9

7.7

22.6

4.6

19.0

39.0

20.5

Presenceof needrated by

patient (%)

8.7

46.2

50.3

14.9

35.9

28.2

68.2

42.1

33.3

9.2

5.1

8.2

2.6

39.5

23.6

14.4

4.6

16.9

3.6

18.5

37.4

22.1

Patients forwhom staffdetect moreneeds thanpatients (%)

1.1

7.9

8.9

13.2

20.5

4.2

27.4

11.1

4.7

5.3

2.1

7.9

6.8

18.9

5.8

2.6

4.2

4.2

1.6

2.6

4.7

4.7

Patients forwhom

patientsdetect moreneeds than

staff (%)

1.1

2.6

2.6

0

2.6

8.9

0.5

6.8

6.8

1.6

1.6

0

0

1.1

3.2

1.6

1.6

1.6

0.5

2.1

2.6

5.8

Agreementon the

presence ofneed (kappa

value)

0.84

0.80

0.75

0.59

0.53

0.66

0.17

0.61

0.73

0.64

0.52

0.62

0.39

0.59

0.71

0.77

0.46

0.80

0.74

0.85

0.79

0.65

Presence ofunmet need

rated bystaff (%)

2.1

3.1

4.1

4.1

29.2

1.5

13.3

2.1

7.2

4.2

3.6

3.1

4.7

38.5

17.4

16.3

2.6

12

3.6

5.6

5.6

7.6

Presence ofunmet need

rated bypatient (%)

2.6

2.1

3.2

0.5

13.7

2.6

8.4

8

11.6

4.2

1.1

0.5

1.1

22.8

17.3

13.8

1.6

10.1

2.1

3.7

4.3

10.9

Patients forwhom staffdetect moreunmet needs

thanpatients (%)

0

1.6

1.6

3.7

17.4

0.5

6.3

1.6

1.6

1.6

2.5

2.5

3.2

17.4

3.7

2.5

1.6

3.2

1.6

2.5

2.5

2.1

Patients forwhom patients

detect moreunmet needsthan staff (%)

0.5

0.5

0.5

0

2.1

1.6

1.6

6.8

6.3

1.6

0.5

0

0

2.5

3.7

1.6

1.1

1.6

0

0.5

1.1

4.2

Agreementon the

presence ofunmet need

(kappavalue)

0.89

0.59

0.70

0.21

0.44

0.49

0.59

0.08

0.53

0.61

0.24

0.30

0.39

0.53

0.71

0.76

0.27

0.75

0.72

0.65

0.57

0.59

S'

I"£tovo

I

9


Dow

nloaded from



was concentrated in daytime activities, company, and inti-mate relationship for both staff and patients. Again, staffdetected more unmet needs than patients, but patientsdetected more unmet needs in psychological distress,physical health, information, accommodation, and bene-fits.

Agreement of rating between patients and staff inpresence of needs was almost perfect (kappa > 0.80) in theareas of accommodation, food, education, and transport.Substantial agreement (kappa between 0.6 and 0.8) wasfound in money, sexual expression, house upkeep, tele-phone, intimate relationship, psychological distress, bene-fits, physical health, risk to self, information, and alcohol.The areas of moderate agreement were self-care, company,daytime activities, child care, and risk to others. Finally,fair agreement was found in drugs and psychotic symp-toms. The agreement on the presence of unmet needs isworse, probably partly because of the low frequency of thepresence of unmet needs. Agreement was low (kappa val-ues lower than 0.4) in self-care, information, risk to others,alcohol, drugs, and child care.

The analysis of the discrepancies showed that formost needs, the percentage of needs or unmet needsdetected by patients but not by staff was low (table 2).Only for physical health, information, psychological dis-tress, intimate relationship, and benefits did patients detectmore needs than staff.

The second objective of our study was to know wherepatients receive help for covering their needs. Table 3shows the percentage of people who receive help and whoprovides that help. Included are not only people with anunmet need but people with a nonsevere need. That is whysome patients in this category are not receiving any formalor informal help. We show the results for only patientsbecause the results for staff are in general similar; we pre-fer to show the results for patients because they describeperceived help. We found that patients receive more infor-mal than formal help: 75 percent of the people with a metneed are receiving informal help, whereas less than 50 per-cent are receiving formal help. Many people with unmetneeds also receive formal and informal help (data notshown), but this help is not sufficient to cover the need.

Table 3. Percentage of informal and formal help received by patients with met needs

Accommodation

Food

House upkeep

Self-care

Daytime activities

Physical health

Psychotic symptoms

Information


Risk to self

Risk to others

Alcohol

Drugs

Companionship


Sexual expression

Child care

Education

Telephone

Transport

Money

Benefits

Patients, n (%)

11 (5.6)

86(44.1)

92 (47.2)

28(14.4)

49(25.1)

50 (25.6)

117(60.0)

67 (34.4)

43(22.1)

10(5.1)

8(4.1)

15(7.7)

3(1.5)

33(16.9)

15(7.7)

6(3.1)

6(3.1)

20(10.3)

3(1.5)

29 (14.9)

65 (33.3)

21 (10.8)

Receivinginformal help (%)

45

94

92

96

86

56

80

54

65

60

63

93

100

76

67

50

67

75

100

55

95

90

Receivingformal help (%)

55

10

5

11

51

70

94

87

84

80

50

80

100

58

20

17

33

35

0

34

3

90

205


Dow

nloaded from


Schizophrenia Bulletin, Vol. 29, No. 2, 2003 S. Ochoa et al.

When analyzing whether people with met needs receivemore help than people with unmet needs, we found thatpeople with met needs tend to receive more help than peo-ple with unmet needs. Patients with met needs in daytimeactivities, information, and intimate relationship receivemore informal help and people with met needs in informa-tion, company, and social benefits receive more formalhelp than people with unmet needs in those categories (allp values below 0.05 or 0.01).

Multiple linear regression was used to determine theassociation between the sociodemographic and clinicalvariables and total number of needs and unmet needs. Fourdifferent models were created, two for staff-determinedneeds (met and unmet) and two for patient-determinedneeds. None of the sociodemographic factors was associ-ated with number of needs. In general, needs were weaklyassociated with the clinical correlates and QOL (R2 lowerthan 0.3 in all models). Correlates of number of needs dif-fer between staff and patients, with patients giving moreweight to disability and QOL and staff to severity ofsymptoms (table 4).

Twenty-two multiple logistic regression models werecreated to determine the influence of sociodemographic,clinical, and social variables on the presence or absence ofeach need (table 5). Men have more problems in the areasof food and house upkeep and women in the areas of trans-port and benefits. Marital status has a significant role inonly daytime activities for unmarried people. Youngerpeople have more problems in the areas of house upkeepand daytime activities and older people in accommoda-tion. People with a longer duration of illness have moreproblems in psychotic symptoms and risk to others, butless in transport. Psychopathology is associated with pres-ence of need in psychological distress, risk to others, riskto self, education, telephone, and money. A worse GAFscore is related to the presence of need in food, houseupkeep, and daytime activities and with absence of need inpsychotic symptoms. Worse QOL is related to the pres-

ence of need in accommodation, company, intimate rela-tionship, daytime activities, and sexual expression.

DiscussionWe have found that people with schizophrenia who live inthe community in Barcelona have a mean number of needsaround 6, with one-quarter of those needs unmet. Thesefigures are similar to those found in England and in Nordiccountries (Slade et al. 1998; Hansson et al. 2001) using theCAN interview but higher than the numbers found in Italy(Lasalvia et al. 2000). In a European study (McCrone et al.2001), the number of needs found in Santander (Spanisharea) is about 5. In analyzing the differences, we need totake into account that two of the studies (Slade et al. 1998;Lasalvia et al. 2000) included patients with psychosisother than schizophrenia. However, the CAN is more aninventory of needs than a questionnaire for which we canobtain a global rating. Although the total number of needsand unmet needs seem to be useful measures because theyindicate the amount of service requirements, the hetero-geneity of the needs contained in the questionnaire impliesthat analysis should be focused on each individual need.

Most of the needs patients have are met. In the caseswhere the need is unmet, more than half of the people arereceiving some kind of help (informal or formal). Whythen do these patients who are receiving help still have anunmet need? It could be because services or informal care-givers do not provide enough help to overcome the need orthe patients are not willing to receive all the help that isavailable. Another possible reason why help might notchange an unmet into a met need is that the need isunmeetable, irrespective of help given. The characteristicsof the CAN do not allow us to differentiate between theoptions.

The most frequent need that we have found is in thearea of psychotic symptoms, which makes sense becausewe have studied the situation of outpatients with schizo-

Table 4. Influence of severity of symptoms, disability, and quality of life on total number of needs andunmet needs

PANSS

GAF

DAS (global)

QOL

fl2

Number

Patient

0.04(0.01,—

0.11 (0.01,—

0.11

0.08)

0.21)

of Needs

0.05

-0.06

—

—

0.25

Staff

(0.01,0.09)

(-0.09, -0.02)

Number of

Patient

—

—

—

-0.044 (-0.02, -0.06)

0.10

Unmet Needs

Staff

0.03(0.01,

—

0.11 (0.04,

—

0.15

0.06)

0.18)

Note.—DAS = Disability Assessment Schedule; GAF = Global Assessment of Functioning Scale; PANSS = Positive and Negative Syn-drome Scale; QOL = Quality of Life questionnaire. Numbers shown are coefficients in a linear regression analysis.

206


Dow

nloaded from


toO

Table 5. Multiple logistic regression: Presence or absence of each need as assessed by the patient as a function of sociodemo- 2?graphic, clinical, and functioning predictor variables |

5?Gender Marital Length of General Negative Positive 5

status Age illness PANSS PANSS PANSS GAF DAS QOL £

0.46*** 0.95* 1

0.95**** |

1.43* 0.95*** I

1.20*** «

0.60* 0.55**** 0.97* 0.97*

0.97* 1.03*

1.02*

1.06**

0.90* 1.03*

0.97*

0.97*

0.96*

1.04*

1.11*

1.04* ^

1.05*** j |

|

AccommodationFood

House upkeep

Self-care

Daytime activities

Physical health

Psychotic symptoms

Information


Risk to self

Risk to others

Alcohol

Drugs

Companionship


Sexual expression

Child care

Education

Telephone

Transport

Money

Benefits

1.70***

1.61**

0.57**

0.65*

Note.—DAS = Disability Assessment Schedule; GAF = Global Assessment of Functioning Scale; PANSS = Positive and Negative Syndrome Scale; QOL = Quality of Life question-naire. Numbers shown are coefficients of the model.* p < 0.05; ** p < 0.01 ; ' " p< 0.005; **** p < 0.001


Dow

nloaded from



phrenia. Besides psychotic symptoms, the most commonneeds found are food, house upkeep, daytime activities,and company. This is in accordance with the work of otherinvestigators, who reported that company, daytime activi-ties, and psychotic symptoms were the most commonneeds (Hansson et al. 1995, 2001; Wiersma et al. 1998).However, these other studies have also detected psycho-logical distress, physical health, and information to becommon needs. Other Spanish studies also found psy-chotic symptoms, daytime activities, and company to bethe most common needs (Rosales 1999; McCrone et al.2001).

In spite of the development of psychiatric communityservices, patients receive more informal than formal help.Except for the areas where clinical treatment and outpa-tient services directed to them are very clearly identified(psychotic symptoms, physical health, information, risk toself, accommodation, and psychological distress), infor-mal help is more common than formal help. The deinstitu-tionalization process relies more on support given by thefamily and the social network of the patient than on sup-port from the health services (Denker and Denker 1994).Formal care services (both mental health care and socialservices) have a low impact on needs related to basicactivities of daily living and social and family relations.Patients may rely on their family and social networks tocover their deficits in these areas. Maybe that is why theseareas are highly correlated with family burden (Maglianoet al. 2000). Mental and social services should addressthese necessities if family QOL is to be improved.

We have found that many people with needs andunmet needs do not receive help from formal services. Thefirst interpretation we have made is that services are notappropriate to cover these needs, but it could also be truethat services are offered to patients but that patients choosenot to receive that help. This, however, would also meanthat services are not tailored enough to patients' prefer-ences, because people who could benefit from them prefernot to use them.

Number of needs and the presence of each need areweakly related to the sociodemographic and clinical vari-ables. People with more severe clinical symptoms andhigher disability have more unmet needs. In accordancewith Slade et al. (1999), patients with a worse QOL havemore unmet needs. However, clinical symptoms, globalfunctioning, QOL, and disability as assessed by thePANSS, the GAF, the QOL, and the DAS predict onlybetween 10 and 25 percent of the variance of number ofneeds. Measurement variance probably has a low influ-ence on this finding: correlation coefficients between thePANSS, the GAF, and the DAS are high (p < 0.001). Whenevaluating needs we are assessing something that is onlypartially related to the clinical status and disability of the

patient. This finding reinforces the need to specificallyevaluate patient needs, because to limit the evaluation tofunctioning, disability, or clinical variables is clearlyinsufficient when deciding the services a patient mayrequire.

Staff detect more needs than patients do. Agreementbetween them was generally fair when evaluating thepresence of needs and substantially lower in the evalua-tion of unmet needs. Slade et al. (1998) suggested thatagreement is better when there is a service that coversthat particular need. Our findings do not support thishypothesis. For example, although a specific serviceaddresses problems with use of drugs or psychotic symp-toms, agreement between staff and patient detection ofneeds was not good. Awareness of illness and negativeand cognitive symptoms may be causing this discrep-ancy. Although administration of the CAN to bothpatients and staff provides the most comprehensive eval-uation, it is also time-consuming. Taking into account theagreement between staff and patients, and the fact thatstaff detect more needs, the CAN may not need to beadministered to both of them; administration to staff onlycan be sufficient. In this case, the patient should evaluatehis or her needs in physical health, psychological dis-tress, and benefits because these are the areas wherepatients detected more needs than staff. With this short-ened administration one could save time and make theinstrument more appropriate for day-to-day practice.

When interpreting the results of the study, we need totake into account that our sample is representative of thepatients that receive outpatient treatment for schizophreniain the public sector. We have not included in the study peo-ple who are not receiving treatment, people who arereceiving it in only the private sector, and patients who areliving in long-stay units (who should have a higher num-ber of needs). To be informed about the representativenessof the sample, we should consider that the 6-month treat-ment prevalence for schizophrenia in the MHCCs includedin the study was 0.28 percent (a total of 1,223 patients in acatchment area that included 439,300 adults), which issimilar to the figures in other countries (Thornicroft et al.1993).

The analysis of the needs of the patients has use notonly to design treatment plans for individual patients butalso to study the limitations of mental health care services.Many questions remain unanswered regarding how needsare covered by formal services and informal help and howthis treatment or help influences the course and the pres-ence of needs. Because the study we present has a cross-sectional design, the findings can be interpreted as onlycorrelations between variables, and no causal inferencesshould be made. Only followup studies with representativesamples of patients will be able to clarify this issue.

208


Dow

nloaded from



The study has several clinical implications:

1. The evaluation of needs and unmet needs is necessaryand complementary to clinical evaluation when design-ing treatment plans for people with schizophrenia.

2. People with schizophrenia receive more informal thanformal care to cover their needs.

3. For most of the needs, staff evaluation may be suffi-cient, but for some needs patient participation in theevaluation is required.

There are also several limitations to the study:

1. The cross-sectional design of the study determines thatno causal relationship can be established from the find-ings.

2. The CAN is more an inventory of needs than an instru-ment to measure overall level of need.

3. We studied only patients who had been in outpatientcare.

ReferencesAmerican Psychiatric Association. DSM-IV: Manualdiagnostico y estadistico de los trastornos mentales.Barcelona, Spain: Masson, 1995.

Andrew, G., and Henderson, S. Unmet Need in Psychiatry.London, U.K.: Cambridge University Press, 2000.

Baker, F., and Intagliata, J. Quality of life in the evalua-tion of community support systems. Evaluation andProgram Planning, 5:69-79, 1982.

Bobes, J.; Saiz, P.A.; and Gonzalez, M.P. Calidad de vidaen las esquizofrenias. Barcelona, Spain: JR Prous, 1995.

Brewin, C.R., and Wing, J.K. Principles and practice ofmeasuring needs in the long-term mentally ill: The Needsfor Care Assessment. Psychological Medicine,17:971-981, 1987.

Denker, S.J., and Denker, K. Does community reduce theneed for psychiatric care? An introduction to theCamberwell High Contact Survey. PsychologicalMedicine, 18:443^146, 1994.

Department of Health Social Services Inspectorate. CareManagement and Assessment: Practitioners' Guide.London, England: HMSO, 1991.

Endicott, J.; Spitzer, R.L.; Fleiss, J.L.; and Cohen, J. TheGlobal Assessment Scale: A procedure for measuringoverall severity of psychiatric disturbance. Archives ofGeneral Psychiatry, 33:766-771, 1976.

Hansson, L.; Bjorkman, T; and Stevenson, B. The assess-ment of needs in psychiatric patients: Interrater reliabilityof the Swedish version of Camberwell Assessment of

Need instrument and results from a cross-sectional study.Acta Psychiatrica Scandinavica, 92:285-293, 1995.

Hansson, L.; Viding, H.; Mackeprang, T.; Sourander, A.;Werdelin, G.; Bentsoon-Tops, A.; Bjamason, O.; Dybbro,J.; Nilsson, L.; Sandlund, M.; Sorgaard, K.; andMiddelboe, T. Comparison on key worker and patientassessment of needs in schizophrenic patients living in thecommunity: A Nordic multicentre study. ActaPsychiatrica Scandinavica, 103:45-51, 2001.

Haro, J.M.; Salvador-Carulla, L.; Cabas6s, J.; Madoz, V.;Vazquez-Barquero, J.L.; and the PSICOST group. Mentalhealth services utilisation and costs of patients with schiz-ophrenia in three areas of Spain. British Journal ofPsychiatry, 173:334-340,1998.

Kay, S.R.; Opler, L.A.; and Fiszbein, A. The Positive andNegative Syndrome Scale (PANSS): Rating manual. Socialand Behavioural Sciences Documents, 17:28-29,1986.

Landis, J., and Koch, G. The measurement of observeragreement for categorical data. Biometrics, 33:159-174,1977.

Lasalvia, A.; Ruggeri, M.; Mazzi, M.A.; and Dall'Agnola,R.B. The perception of needs for care in staff and patientsin community-based mental health services: The South-Verona Outcome Project 3. Acta PsychiatricaScandinavica, 102:366-375, 2000.

Magliano, L.; Fadden, G.; Economan, M.; Held, T.;Xavier, M ; Guarnen, M; Malagone, C ; Marasco, C; andMaj, M. Family burden and coping strategies in schizo-phrenia: 1 year follow up data from the BIOMED I study.Social Psychiatry and Psychiatric Epidemiology,35(3): 109-115, 2000.

Marshall, M. How should we measure need? Philosophy,Psychiatry, and Psychology, 1:27-36, 1994.

McCrone, P.; Leese, M.; Thornicroft, G.; Schene, A.H.;Knudsen, H.C.; Vdzquez-Barquero, J.L.; Lasalvia, A.;Padfield, S.; White, I.R.; Griffiths, G.; and the EPSILONStudy Group. Reliability of the Camberwell Assessmentof Need: European version. EPSILON Study 6. BritishJournal of Psychiatry, 177(39):S34-S40, 2000.

McCrone, P.; Leese, M.; Thornicroft, G.; Schene, A.;Knudsen, H.C.; Vdzquez-Barquero, J.L.; Tansella, M.;Becker, X; and the EPSILON study group. A comparisonof needs of patients with schizophrenia in five Europeancountries: The EPSILON study. Acta PsychiatricaScandinavica, 103:370-379,2001.

Netten, A., and Beecham, J. Costing Community Care.Aldershot, England: Ashgate, 1993.

Norusis, M. SPSS para Windows: Version 6.0 Guia delusuario. Madrid, Spain: Departmento de Estadistica yEconomia, 1993.

209


Dow

nloaded from



Peralta, V., and Cuesta, MJ. Validacion de la escala de lossindromes positivo y negativo en una muestra deesquizofrenicos espanoles. Adas Luso-Espaholas deNeurologia, Psiquiatria y Ciencias Afines, 1994.

Phelan, M.; Slade, M.; Thornicroft, G.; Dunn, G.;Holloway, F.; Wykes, T.; Strathdee, G.; Loftus, L.;McCrone, P.; and Hayward, P. The CamberwellAssessment of Need (CAN): The validity and reliabilityof an instrument to assess the needs of people with severemental illness. British Journal of Psychiatry,167:589-595, 1995.

Rosales, C. "Estudio de las personas con trastornosesquizofrenicos." Tesis doctoral, Granada University ofPsychiatry, 1999.

Sartorius, N.; Jablenski, A.; Korten, A.; Ernberg, G.;Anker, M.; Cooper, J.E.; and Day, R. Early manifestationsand first-contact incidence of schizophrenia in differentcultures: A preliminary report on the initial evaluation ofthe WHO collaborative study on determinants of outcomeof severe mental disorders. Psychological Medicine,16:909-928, 1986.

Slade, M. Needs assessment: Involvement of staff andusers will help to meet needs. British Journal ofPsychiatry, 165:293-296, 1994.

Slade, M.; Leese, M.; Taylor, R.; and Thornicroft, G. Theassociation between needs and quality of life in an epi-demiologically representative sample of people with psy-chosis. Acta Psychiatrica Scandinavica, 100:149—157,1999.

Slade, M.; Phelan, M.; and Thornicroft, G. A comparisionof needs assessed by staff and by an epidemiologicallyrepresentative sample of patients with psychosis.Psychological Medicine, 28:543-550, 1998.

Slade, M.; Phelan, M; Thornicroft, G.; and Parkman, S.The Camberwell Assessment of Need (CAN):Comparision of assessments by staff and patients of theneeds of the severely mentally ill. Social Psychiatry andPsychiatric Epidemiology, 31:109-113, 1996.

Stevens, A., and Gabbay, J. Needs assessment. HealthTrends, 23(l):20-23, 1991.

Thornicroft, G. The uses and limitations of mental healthservice need indexes. British Journal of Psychiatry,169(6):792, 1996.

Thornicroft, G.; Bisoffi, G.; De Salvia, D.; and Tansella,M. Urban-rural differences in the associations betweensocial deprivation and psychiatric service utilization inschizophrenia and all diagnoses: A case-register study inNorthern Italy. Psychological Medicine, 23:487-496,1993.

Wiersma, D.; Nienhuis, F.J.; Giel, R.; and Slooff, C.J.Stability and change in needs of patients with schizo-phrenic disorders: A 15- and 17-year followup from firstonset of psychosis and a comparision between objectiveand subjective assessment of needs for care. SocialPsychiatry and Psychiatric Epidemiology, 33:49-56,1998.

World Health Organization. ICD-10: InternationalClassification of Diseases, Mental and BehaviourDisorders. 10th ed. Geneva, Switzerland, 1992.

Acknowledgments

This project has received the financial help of the SpanishFondo de Investigaciones Sanitarias (FIS 97/1275) and anunrestricted educational grant from Eli Lilly and Co.(ESQ00196).

The Authors

S. Ochoa, M.D., is Psychologist, Sant Joan de D6u-SSM, Sant Boi, Spain. J.M. Haro, M.D., Ph.D., is Psy-chiatrist, Sant Joan de D6u-SSM, Sant Boi, Spain. J.Autonell, M.D., is Psychiatrist, Sant Joan de Deu-SSM,Sant Boi, Spain. A. Pendas, M.D., is at Sant Joan deDeu-SSM, Cerdanyola, Spain. F. Teba, M.D., is at SantJoan de D6u-SSM, Gava, Spain. M. Marquez, M.D.,Ph.D., is at Sant Joan de Deu-SSM, Cerdanyola, Spain.The NEDES group is a multidisciplinary group ofresearchers that includes S. Araya, P. Asensio, J.Autonell, A. Benito, E. Busquets, C. Carmona, P.Casacuberta, M. Castro, N. Diaz, M. Dolz, A. Foix, J.M.Giralt, A. Gost, J.M. Haro, M. Marquez, F. Martinez, R.Martinez, J. Miguel, M.C. Negredo, S. Ochoa, Y. Oso-rio, E. Paniego, L. Pantinat, A. Pendas, C. Pujol, J.Quilez, J. Ramon, M.J. Rodriguez, B. Sanchez, A. Soler,F. Teba, N. Tous, J. Usall, M. Valdelomar, J. Vaquer, E.Vicens, and M. Zamora.

210


Dow

nloaded from
