Jpd223$$4u

Journal of Personality Disorders, 22(3), 259–268, 2008 2008 The Guilford Press THE BURDEN OF DISEASE IN PERSONALITY
DISORDERS: DIAGNOSIS-SPECIFIC QUALITY
OF LIFE
Djøra I. Soeteman, MS, Roel Verheul, PhD,and Jan J.V. Busschbach, PhD A generic quality of life measure was used to investigate the burden ofdisease in a large sample of patients with personality disorders. The1,708 subjects included in this study were recruited from six differentmental health care institutes in the Netherlands. The burden of diseasewas measured using the EuroQol EQ-5D. Personality disorders were di-agnosed using the Structured Interview for DSM-IV Personality (SIDP-IV).
The mean EQ-5D index value was 0.56. Primarily the total number ofpersonality disorder diagnoses rather than the specific type determinedthe quality of life. Notably borderline personality disorder was not asso-ciated with the highest burden. The findings indicate that patients withpersonality disorders experience a high burden of disease, comparableto that of severe somatic illnesses. The results call into question theprimary focus in literature on borderline personality disorder. The cur-rent study yields a strong argument in favor of reimbursing (effective)treatments for this patient population.
Personality disorders are known to be associated with significant impair-ment in social, occupational, and other important areas of functioning.
Several studies indicate poorer social and interpersonal functioning, andpoorer occupational functioning, satisfaction, and achievement amongpatients with personality disorders as compared with others (Lim, Sander-son, & Andrews, 2000; Oltmanns, Melley, & Turkheimer, 2002). On mea-sures of global functioning, most studies have shown significant functionalimpairments for patients with personality disorders (Abrams, Alexopoulos, From Viersprong Institute for Studies on Personality Disorders (VISPD; D. I. S., R. V.,J. J. V. B.); University of Amsterdam, Department of Clinical Psychology (R. V.); and ErasmusMedical Center Rotterdam, Department of Medical Psychology and Psychotherapy (D. I. S.,J. J. V. B.).
The Study on the Cost-Effectiveness of Personality disorder TREatment (SCEPTRE) is beingconducted with the participation of six mental health care institutes in the Netherlands, i.e.,Center of Psychotherapy De Viersprong, Halsteren; Altrecht, Utrecht; Zaans Medisch Cen-trum, Zaandam; Center of Psychotherapy De Gelderse Roos, Lunteren; GGZWNB, Bergen opZoom; Center of Psychotherapy Mentrum, Amsterdam.
Address correspondence to Djøra Soeteman, Viersprong Institute for Studies on PersonalityDisorders (VISPD), P.O. Box 7, 4660 AA Halsteren, The Netherlands, +31 164 632200; E-mail: [email protected] Spielman, Klausner, & Kakuma, 2001; Hueston, Mainous, & Schilling,1996; Johnson et al., 1996). A study by Skodol and colleagues (2002), forinstance, compared the psychosocial functioning in patients with person-ality disorder (schizotypal, borderline, avoidant, and obsessive-compul-sive) with major depressive disorder. They found patients with schizotypaland borderline personality disorder to be more impaired than patients withobsessive-compulsive personality disorder or major depressive disorder.
This impairment is remarkable because impairment in major depressivedisorder has been found to be comparable to that of chronic medical ill-nesses such as diabetes and arthritis (Hayes, Wells, Sherbourne, Rogers,& Spritzer, 1995; Wells et al., 1989). A study of outpatients by Nakao,Gunderson, and Phillips (1992), showed that patients with any personalitydisorder are more functionally impaired (GAF-scores) than those withouta personality disorder.
All these investigations studied the “functioning” in personality disorder patients. However, in contemporary research, more and more emphasis isput on the “subjective quality of life” of these patients. This paradigm shiftis reinforced by a number of studies showing that it is the patients’ subjec-tive well being, rather than objective medical condition, that determinestheir treatment-seeking behavior, their compliance and their evaluation oftreatment (Hunt, & McKenna, 1993). Furthermore, quality of life has be-come an important outcome in cost-effectiveness analysis (Drummond,Sculpher, Torrance, O’Brien, & Stoddart, 2005). As a consequence, theinterest of psychologists and psychiatrists is no longer limited to symp-tom-focused outcome assessment, as they have become aware of the im-portance of quality of life measures in their clinical outcome measures.
In a recent Norwegian study, the quality of life of 72 patients with per- sonality disorders in a psychiatric outpatient clinic was examined (Narud,Mykletun, & Dahl, 2005). The investigators used the multi-dimensionalShort Form 36 (SF-36), a standardized generic measure, to assess thequality of life. The main finding of this study was that personality disorderpatients treated in a psychiatric outpatient clinic had a significantly lowerquality of life, on both the physical and mental SF-36 dimensions, than anage- and gender-adjusted general population sample. Furthermore, in agroup of 1651 inpatients with complex personality problems and personal-ity disorders, Soeteman, Timman, Trijsburg, Verheul, and Busschbach(2005) found a severe impairment in quality of life (EuroQol EQ-5D indexscore of .54). They compared the quality of life in this mental conditionwith those in severe somatic illnesses such as Parkinson’s disease (EQ-5Dindex score = .58) and rheumatic disease (EQ-5D index score = .53). Bothstudies described above have to be considered explorative studies of qual-ity of life in personality disorders. The Norwegian study used a small sam-ple of 72 psychiatric outpatients, and has therefore a limited externalvalidity. Soeteman and colleagues’ sample size was substantial, but nostandardized Axis II diagnoses were available; thus the results have lim-ited internal validity.
The aim of this study is to investigate the relation of the burden of dis- ease in terms of quality of life with the 14 DSM-IV personality disordersusing a generic quality of life questionnaire, i.e., the EuroQol EQ-5D. Sucha generic instrument can measure the burden of disease regardless of pa-tients’ diagnoses and can therefore be used to compare the burden of dis-ease in patients with personality disorders with patients with other medicalconditions, for example severe somatic illnesses. Moreover, the different di-mensions of quality of life in the EQ-5D are combined into one weightedscore, thereby yielding unambiguous comparisons. Note that in this investi-gation we assume an inverse relation between quality of life and burden ofdisease; this assumption is also made in the Global Burden of Disease proj-ect of the WHO (Ustun, Ayuso-Mateos, Chatterji, Mathers, & Murray, 2004).
METHOD
PARTICIPANTS
Participants were recruited from a consecutive series of admissions to sixmental health care institutes in the Netherlands offering outpatient, dayhospital, and/or inpatient psychotherapy for adult patients with personal-ity pathology and/or personality disorders. As part of the standard admis-sion procedure, all applicants performed a routinely distributed assess-ment battery including self-report questionnaires in order to measurepsychopathology, personality, functional impairments, and treatment his-tory, and a semi-structured interview for diagnosing personality disorders.
When the administration of the questionnaires forms part of the routinelyadministered clinical intake procedure and does not involve additionalrisks or load, informed consent is not mandatory under Dutch law. Forthis reason, informed consent was only asked if the patient participated inany further follow-up investigations.
From March 2003 to March 2006, 2,540 individuals have been regis- tered as admissions to the six mental health care institutes. Of these pa-tients, 462 (18.2%) did not start and 272 (10.7%) did not complete theformal admission procedure. Of the remaining 1,806 patients, 46 were ex-cluded due to clear signs of unreliable data in the interview and/or ques-tionnaires (2.3%) or due to serious intellectual impairment (0.3%). TheEQ-5D was missing or incomplete for 52 patients, leaving 1,708 patientsfor the current study sample, i.e., 94.6% of those who completed the for-mal assessment procedure.
Of these patients, 35.4% were male. The mean age was 33.7 years (SD 9.9, range 18–67). Of these, 65.6% were unmarried, 22.0% married, and12.4% were divorced or widowed. No differences with respect to gender,age, and educational level were found between those admissions that wereincluded as compared to those who were excluded from the sample.
The quality of life was measured using the EuroQol EQ-5D (Brooks, Rabin,& de Charro, 2003). The descriptive system of the EQ-5D records quality of life in 5 dimensions: mobility (walking about), self-care (washing anddressing oneself), usual activities (e.g., work, study, housework, family, orleisure activities), pain/discomfort and anxiety/depression. Each dimen-sion is divided into 3 response levels: no problems, some or moderateproblems, and extreme problems or unable to. The combination of scoresdefine a total of 243 different possible health states and each of these areweighted to arrive at a single index score between −0.33 (worst imaginablehealth state) and 1.00 (best imaginable health state). The Dutch normscores were used for calculating the mean EQ-5D index values (Lamers,Stalmeier, McDonnell, Krabbe, & Busschbach, 2005).
Personality disorders were measured using the Dutch version of the Structured Interview for DSM-IV Personality (SIDP-IV; Pfohl, Blum, & Zim-merman, 1995; translated by De Jong, Derks, Van Oel, & Rinne, 1996).
This instrument includes the 11 formal DSM-IV-TR Axis II diagnoses (e.g.,schizoid personality disorder) including personality disorder mixed, thetwo DSM-IV-TR appendix diagnoses (depressive and negativistic personal-ity disorder), and—in addition—the DSM-III-R self-defeating personalitydisorder. Interviewers were master-level psychologists, who were trainedthoroughly by one of the authors (RV), and who received monthly boostersessions to avoid drift from the interviewer guidelines. Inter-rater reliabil-ity was computed in 30 videotaped interviews rated by three observer-raters. Percentage agreement ranged from 84% (avoidant PD) to 100%(schizoid; median 95%). Intraclass correlation coefficients (ICC) for thesum of DSM-IV personality disorder traits present (i.e., scores “2” or “3”)ranged from 0.60 (schizotypal) through 0.92 (antisocial; median 0.74).
A multiple regression main effect analysis was conducted, measuring theindependent contribution of the different diagnoses on quality of life. Themajority of patients (54.9%) received at least two personality disorder diag-noses. That is the reason an additional regression was performed to ac-count for possible interactions between diagnoses. Because the number ofpossible interactions between 14 independent variables becomes intracta-ble, the interaction term is represented by a count of the diagnoses given.
Age, gender, and education (socioeconomic status) variables are associ- ated with quality of life and were therefore entered into the multiple regres-sion models (Brooks et al., 2003).
RESULTS
In Table 1, the rank ordering of the quality of life figures is displayed for
the 14 specific DSM-IV personality disorders. Because patients can have
more than one personality disorder, the sum of the number of patients in
TABLE 1. EuroQol EQ-5D Index Scores (Mean and Standard Deviation)
for the 14 DSM-IV Personality Disorders
Analysis
Personality disorder
N
SD
p
1Linear regression analysis: dependent variable quality of life; independent variables cat-egorical diagnoses: having or not having that particular personality disorder the different diagnostic groups is higher than the total number of patientsincluded in this study.
The mean EQ-5D index value for the personality disordered group as a whole was .56 (SD = .27), representing a severe burden of disease. Notethat the mean EQ-5D index scores for almost all of the specific diagnosticgroups (except for PD mixed and schizotypal personality disorder) in Table1 appear to be lower than the mean EQ-5D index score for the total groupof patients with at least one disorder (.56). This is possible because pa-tients with a large number of diagnoses, and concordantly a low quality oflife (see also Table 2), are represented in an equally large number of diag-nostic groups. As a consequence of their poor quality of life, they “lower”the mean EQ-5D index scores of all of these groups.
In the present sample, depressive (32.0%), avoidant (28.5%), obsessive- compulsive (20.8%), and borderline personality disorder (20.8%) were themost frequently diagnosed disorders. Schizotypal (0.9%) and schizoid per-sonality disorder (1.1%) were the least frequently diagnosed disorders. Inabout one-fifth of the total group of patients no personality disorder couldbe diagnosed.
When studying the main effects of the specific personality disorders in a linear regression analysis, six out of 14 appeared significant (p < 0.05),indicating that having or not having that specific disorder has a significanteffect on the quality of life in this sample. These six disorders are border-line, narcissistic, obsessive-compulsive, depressive, negativistic personal-ity disorder, and personality disorder mixed.
Table 2 shows that the quality of life is inversely associated with the number of personality disorders diagnosed. As could be predicted from the TABLE 2. EuroQol EQ-5D Index Scores (Mean
and Standard Deviation) for Increasing Number
of Personality Disorder Diagnoses
Number of PDs
N
index score
SD
lower means for specific diagnoses compared to the overall mean in Table1, the number of personality disorders has a large effect on quality of life(p = 0.000). When controlling for the number of disorders in the regressionanalysis, only depressive personality disorder maintains a unique statisti-cally significant effect on quality of life (p = 0.03).
DISCUSSION
Personality disorders are associated with a severe impairment in quality
of life. The overall EQ-5D index value of .56 suggests that the quality of
life experienced by patients with personality disorders can be compared
to the quality of life in, for instance, rheumatic disease, lung cancer, or
Parkinson’s disease with EQ-5D index scores of .53, .58, and .58, respec-
tively (Siderowf, Ravina, & Glick, 2002; Trippoli, Vaiani, Lucioni, & Mes-
sori, 2001; Wolfe & Hawley, 1997). The burden of having a personality
disorder seems even higher than in patients with type II diabetes (EQ-5D
score of .69), schizophrenia outpatients treated with neuroleptics (.73),
and HIV infected patients (.77; Dernovsek, Prevolnik Rupel, Rebolj, &
Tavcar, 2001; Koopmanschap, 2002; Stavem, Frøland, & Hellum, 2005).
The burden is only found to be higher in major depressive disorder (.33)
and patients with renal failure on heamodialysis (.44; Lee, Morgan, Con-
way, & Currie, 2005; Sapin, Fantino, Nowicki, & Kind, 2004). It can be
concluded that patients who are in search for treatment for their personal-
ity disorders experience a high burden of disease, as compared to other
populations with severe somatic illnesses.
Borderline, narcissistic, obsessive-compulsive, depressive, negativistic personality disorder and personality disorder mixed appear to have a sig-nificant effect on the quality of life. However, when the total number ofpersonality disorders diagnosed is taken into account and included in theanalysis, the latter appears the most important predictor of quality of life,leaving only the depressive personality disorder with an additional effect.
These findings seem to imply that in patients with borderline, narcissistic,obsessive-compulsive, negativistic personality disorder, and personalitydisorder mixed the comorbidity of other Axis II disorders rather than thespecific diagnosis caused the quality of life to be more impaired. This con- clusion is in line with the results of a study by Jackson and Burgess(2004), in which no significant differences could be found on the SF-12Mental Summary Scale between eight ICD-10 personality disorders in anAustralian community sample of 10,641 participants. Consistently, theauthors mentioned that the addition of other comorbid personality disor-ders to the specific personality disorders led to sizeable increases in theodds ratios for all personality disorder types on the disability measures.
Similar results were found by Nakao et al. (1992), who showed a strongpositive relationship between the total number of Axis II criteria met andthe severity of the functional impairment as measured with the Global As-sessment of Functioning Scale (GAF score). Jackson and Burgess (2000)also found an increasing disability according to the SF-12 Mental Sum-mary Subscale with an increasing number of personality disorders. Theonly exception seems the study by Narud et al. (2005) who found no wors-ening of quality of life as measured with the SF-36, with increasing num-ber of personality disorders. They attributed this to a small sample size.
The 1,708 participants included in this study were recruited from six different mental health care institutes in the Netherlands specialized inthe psychotherapeutic treatment of personality problems and disorders.
The large number of patients and the different settings can be consideredone of the strengths of this study, as these enhance the external validityof the results. On the other hand, we only sampled patients that were re-ferred to some sort of psychotherapy; therefore the results may not be gen-eralized to all prevalent cases in the community. However, this can not beconsidered legitimate criticism taken into account the aim of the study.
The ultimate purpose of our study is providing an argument for reimburs-ing, in other words providing an answer to the question if an expensivetreatment for this population of patients, based on the necessity of treat-ment (i.e., burden of disease), is justified. Only personality disordered pa-tients who actually search for, or are referred to, some sort of treatmentwill claim money for that treatment, which is paid for by society.
Another limitation of our study is that no standardized diagnoses of co- morbid Axis I psychiatric disorders were available. Note, however, that thislimitation does not jeopardize the main finding of our study, namely thatpatients who seek treatment for their personality disorders experience ahigh burden of disease. In this stage of their disorder, when patients areadmitted to a mental health care facility, it is difficult to find patients with“only” Axis II problems. Isolating the effects of the Axis I disorders wouldbe the same as considering the burden of disease of diabetics without thefoot ulcers or the quality of life of schizophrenic patients without the symp-toms caused by neglect. Moreover, it has been shown that Axis I and AxisII disorders are independently related to disease-specific burden of diseaseparameters (Jackson & Burgess, 2000, 2002; Skodol et al., 2002; Verheulet al., 2000). The independent contribution of Axis I and Axis II pathologyto the burden of disease should be addressed in future research.
The use of generic quality of life measures in mental health research has been criticized (Chisholm, Healey, & Knapp, 1997). According to Chisholm and colleagues, one of the concerns is that the domains of particular im-portance in the measurement of quality of life in people with mental healthproblems are not represented properly in the prominent generic quality oflife measures employing domains of physical mobility, pain, and disability.
They argue that this can lead to an undervaluation of the burden of dis-ease in the mentally ill. Another concern is that mental disorders are per-ceived as more heterogeneous in the course, content, and consequencesover time than somatic disorders, which causes the quality of life in mentaldisorders to have a limited predictability and stability.
However, this study provides evidence that these concerns are not justi- fied for the patient population subject to our investigation. A substantialburden of disease was found by using the generic EuroQol EQ-5D, whichat least indicates that an important part of the problems in this particularpatient group are well captured in the 5 domains of the EQ-5D. Moreover,a similar high burden of disease (EuroQol EQ-5D index score of .54) wasfound in an earlier study of Soeteman et al. (2005) among a large group ofpatients with similar problems, which indicates the robustness of thepresent findings. Additionally, the reliability and validity of quality of lifemeasures have been established in other mental illnesses, such as schizo-phrenia (Pukrop et al., 2003). The present demonstration of the use of theEQ-5D in personality disorders should encourage its use in research,which should help positive funding decisions since it is easier to makecomparative decisions across disease types using generic quality of lifemeasures such as the EQ-5D.
When examining the ranking of the burden of disease in the 14 specific DSM-IV personality disorders in Table 1, it becomes clear that a high bur-den is not necessarily associated with receiving more attention in clinicalresearch. Blashfield and Intoccia (2000) have shown that the only person-ality disorder whose literature was clearly alive and growing was that ofthe borderline personality disorder; a disorder that is positioned in theupper regions of the ranking of quality of life. On the other hand personal-ity disorders that are associated with a higher burden, according to ourranking, have either very small literatures (e.g., dependent, narcissistic,paranoid, passive-aggressive) or literatures with flat or negative growthrates (e.g., dependent, histrionic, paranoid, passive-aggressive, schizoid).
One explanation is that the disorders that cause the greatest societal bur-den (e.g., antisocial) or the greatest burden to clinicians (e.g., borderline)have traditionally attracted most scientific attention. Our findings suggestthat an emphasis on burden from a patient perspective would have leadto completely different choices.
In health care, cost-effectiveness analyses are a well-established deci- sion tool in reimbursement policy. However, a growing body of evidencesuggests that cost-effectiveness alone is not sufficient for rational decisionmaking in this regard. It is found that burden of disease interacts withcost-effectiveness considerations: the higher the burden of disease, themore willing society is to accept a poor cost-effectiveness (Pronk, & Bonsel,2004; Stolk, Brouwer, & Busschbach, 2002). For instance, the cost-effec- tiveness of Viagra is very favorable, but its funding remains in dispute. Onthe other hand, lung-transplantation is known for its unfavorable cost-effectiveness, yet the reimbursement is not a matter of debate. It thusseems that the burden of the patients (or how pitiful their situation seems)also plays a key role in the discussion which treatments to fund. Moreand more existing treatments, which have long been reimbursed withoutproviding any evidence for their cost-effectiveness such as, for instance,psychotherapy, are recently required to demonstrate their efficiency in or-der to free budget for the treatments, which have already shown to be cost-effective. The current study, showing a high burden of disease in patientswith personality disorders, yields a strong argument in favor of reimburs-ing (effective) treatments for this patient population.
REFERENCES
Abrams, R. C., Alexopoulos, G. S., Spielman, L. A., Klausner, E., & Kakuma, T.
chronic general medical illness. Ar- chives of General Psychiatry, 52, 11–19.
Hueston, W. J., Mainous, A. G., & Schilling, patients. American Journal of Geriatric Psychiatry, 9, 67–71.
Blashfield, R. K., & Intoccia, V. (2000).
care. Journal of Family Practice, 42, personality disorders. American Jour- Hunt, S. M., & McKenna, S. P. (1993). Mea- nal of Psychiatry, 157, 472–473.
Brooks, R., Rabin, R., & de Charro, F. (2003).
Quality of Life Assessment: Key issues in the 1990s. (S.R. Walker & R.M. Ros- health status using EQ-5D: A European perspective. London: Kluwer Academic Jackson, H. J., & Burgess, P. M. (2000). Per- Chisholm, D., Healey, A., & Knapp, M.
(1997). QALYs and mental health care.
Social Psychiatry and Psychiatric Epi- Survey of mental health and wellbeing.
demiology, 32, 68–75.
Social Psychiatry and Psychiatric Epi- De Jong, C.A.J., Derks, F.C.H., Van Oel, C. J., demiology, 35, 531–538.
& Rinne, T. (1996). Gestructureerd Inter- Jackson, H. J., & Burgess, P. M. (2002). Per- view voor de DSM-IV persoonlijkheidss- toornissen: SIDP-IV. Sint Oedenrode: Stichting Verslavingszorg, Oost-Brabant.
Survey of mental health and wellbeing.
Dernovsek, M. Z., Prevolnik Rupel, V., Re- Social Psychiatry and Psychiatric Epi- bolj, M., & Tavcar, R. (2001). Quality demiology, 37, 251–260.
Jackson, H. J., & Burgess, P. M. (2004). Per- pot neuroleptics. European Psychiatry, Drummond, M. F., Sculpher, M. J., Torrance, part III. Social Psychiatry and Psychi- G. W., O’Brien, B. J., & Stoddart, G. L.
atric Epidemiology, 39, 765–776.
(2005). Methods for the economic evalu- Johnson, J. G., Williams, J.B.W., Goetz, R. R., ation of health care programmes. Ox- J. D., & Gorman, J. M. (1996). Person- Hayes, R. D., Wells, K. B., Sherbourne, C. D., Rogers, W., & Spritzer, K. (1995). Func- risk of HIV infection. Archives of Gen- eral Psychiatry, 53, 350–357.
Koopmanschap, M. (2002). Coping with type disorder. Health and Quality of Life II diabetes: The patient’s perspective.
Diabetologia, 45, S18–S22.
Siderowf, A., Ravina, B., & Glick, H. A.
Lamers, L. M., Stalmeier, P.F.M., McDonnell, in patients with Parkinson’s disease.
Neurology, 59, 103–108.
Skodol, A. E., Gunderson, J. G., McGlashan, Dutch EQ-5D-tarif (in Dutch). Neder- T. H., Dyck, I. R., Stout, R. L., Bender, lands Tijdschrift voor de Geneeskunde, Lee, A. J., Morgan, C. L., Conway, P., & Cur- sive personality disorder. American Journal of Psychiatry, 159, 276–283.
life for patients with renal failure. Cur- Soeteman, D. I., Timman, R., Trijsburg, R. W., rent Medical Research and Opinion, 21, Verheul, R., & Busschbach, J.J.V.
Lim, D., Sanderson, K., & Andrews, G.
ized units that provide psychotherapy.
Psychiatric Services, 56, 1153–1155.
The Journal of Mental Health Policy Stavem, K., Frøland, S. S., & Hellum, K. B.
and Economics, 3, 139–146.
Nakao, K., Gunderson, J. G., & Phillips, personality disorders. Journal of Per- Quality of Life Research, 14, 971–980.
sonality Disorders, 6, 24–33.
Stolk, E. A., Brouwer, W.B.F., & Bussch- Narud, K., Mykletun, A., & Dahl, A. (2005).
chiatric outpatient clinic. BMC Psychi- Viagra. Health Policy, 59, 53–63.
Trippoli, S., Vaiani, M., Lucioni, C., & Mes- Oltmanns, T. F., Melley, A. H., & Turkheimer, sori, A. (2001). Quality of life and util- cancer. Pharmacoeconomics, 19, 855– non-clinical population. Journal of Per- Ustun, T. B., Ayuso-Mateos, J. L., Chatterji, sonality Disorders, 16, 437–452.
S., Mathers, C., & Murray, C. J. (2004).
Pfohl, B., Blum, N., & Zimmerman, M.
in the year 2000. British Journal of Psychiatry, 184, 386–392.
Verheul, R., Kranzler, H. R., Poling, J., Ten- Pronk, M. H., & Bonsel, G. J. (2004). Out- nen, H., Ball, S., & Rounsaville, B. J.
Axis II disorders in substance abusers.
straints? European Journal of Health Acta Psychiatrica Scandinavica, 101, Economics, 5, 274–277.
Pukrop, R., Schlaak, V., Moller-Leimkuhler, Wells, K. B., Stewart, A., Hayes, R. D., Bur- kotter, J., & Moller, H. J. (2003). Reli- ability and validity of quality of life as- from the medical outcome study. Jour- nal of the American Medical Associa- tients with depression. Psychiatry Re- Wolfe, F., & Hawley, D. J. (1997). Measure- Sapin, C., Fantino, B., Nowicki, M. L., & disorders using the EuroQol. British Journal of Rheumatology, 36, 786–793.

Source: http://www.deviersprong.nl/files/Paginas/2008JPDSoetemanQoL.pdf