SEARCH

SEARCH BY CITATION

Keywords:

  • Mini-International Neuropsychiatric Interview;
  • psychiatric diagnosis;
  • reliability;
  • Structured Clinical Interview for DSM-III-R;
  • structured diagnostic interview;
  • validity

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSIONS
  8. ACKNOWLEDGMENTS
  9. REFERENCES

Abstract  The Mini-International Neuropsychiatric Interview (MINI) is a short, structured diagnostic interview used as a tool to diagnose 16 axis I (Diagnostic and Statistical Manual) DSM-IV disorders and one personality disorder. Its original version was developed by Sheehan and Lecrubier. We translated the MINI into Japanese, and investigated the reliability and validity of the Japanese version of MINI. Eighty-two subjects participated in the validation of the MINI versus the Structured Clinical Interview for DSM-III-R (SCID-P). One hundred and sixty-nine subjects participated in the validation of the MINI versus an expert's professional opinion. Seventy-seven subjects were interviewed by two investigators and subsequently readministered by a third interviewer blind to the results of initial evaluation 1–2 days later. In general, kappa values indicated good or excellent agreement between MINI and SCID-P diagnoses. Kappa values indicated poor agreement between MINI and expert's diagnoses for most diagnoses. Interrater and test–retest reliabilities were good or excellent. The mean durations of the interview were 18.8 min for MINI and 45.4 min for corresponding sections of SCID-P. Overall, the results suggest that the MINI Japanese version succeeds in reliably and validly eliciting symptom criteria used in making DSM-III-R diagnoses, and can be performed in less than half the time required for the SCID-P.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSIONS
  8. ACKNOWLEDGMENTS
  9. REFERENCES

To conduct large-scale epidemiologic studies, or multicenter clinical trials in psychopharmacology, diagnostic structured interviews are used for the screening of homogenous samples. However, investigators often find that the administration of comprehensive diagnostic interviews is time-consuming and expensive. The most widely used structured interviews, namely, the Structured Clinical Interview for Diagnostic and Statistical Manual (DSM)-III-R1-patient version (SCID-P),2,3 Composite International Diagnostic Interview for ICD-10 (CIDI),4,5 Schedules for Clinical Assessment in Neuropsychiatry (SCAN),6 and Schedule for Affective Disorders and Schizophrenia (SADS),7 contain more than 500 questions and require 1–2 h to conduct. Long interviews can be difficult to conduct when patients have trouble concentrating or are agitated. In addition, training  in  the  use  of  precise  wordings  and  algorithmic decisions is long, usually approximately 5 days. Although two short structured diagnostic interviews (the Symptom-Driven Diagnostic System [SDDS]8,9 and the Primary Care Evaluation of Mental Disorders [Prime-MD]10) have been developed, they were designed for primary care and are limited to screening the six most common psychiatric disorders observed in primary care.

The Mini-International Neuropsychiatric Interview (MINI)11–13 was developed by Sheehan and Lecrubier to meet the need for a brief, reliable and valid structured diagnostic interview that screens more disorders than SDDS-PC or Prime-MD, and possibly replaces SCID-P or CIDI in clinical trials. The MINI contains 130 questions and screens 16 axis I DSM-IV14 disorders and one personality disorder. Similar to SCID-P and CIDI, the MINI is organized in diagnostic modules. For most modules, two to four screening questions are used to rule out the diagnosis when answered negatively. Positive responses to screening questions are explored by further investigation of other diagnostic criteria. Excellent interrater and test–retest reliabilities of the English version of MINI,11–13 and moderate validity of MINI versus CIDI11,13 and SCID-P12,13 were reported. The MINI is gaining rapid acceptance internationally, being translated into many languages by 2004. However, non-English versions of the MINI have not yet been reported, hence the cross-cultural validity of MINI is not yet known.

The authors developed the Japanese version of the MINI with the cooperation of the original developers of the tool. The aim of this study was to determine its reliability and validity. We needed an validation study of the MINI in both psychiatric and non-psychiatric patients. However, this is the first validation study of the Japanese version of the MINI, and we need an adequate number of samples for each of 16 major psychiatric disorders. The aim of this study was to determine its reliability and validity in the ‘clinical psychiatry setting’.

METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSIONS
  8. ACKNOWLEDGMENTS
  9. REFERENCES

Translation of the MINI

With the original authors’ permission, our team in the Department of Psychiatry of the School of Medicine, Showa University translated the original English MINI into Japanese. We followed the standard procedure of backtranslation for the cross-cultural adaptation of an original English psychometric instrument.15 Namely, one of the authors (TO) first translated the English MINI into Japanese. This preliminary Japanese MINI was backtranslated into English by two Japanese psychiatrists who had lived in English-speaking countries for 1–5 years. The backtranslated version was then examined against the original scale by Sheehan who pointed out possible discrepancies. TO then corrected the Japanese translation accordingly. This process was repeated three times, until all authors agreed that the original and backtranslated versions matched closely. Throughout this procedure the two men who backtranslated the Japanese MINI into English remained blind to the original English version. The Japanese version of the MINI was published by Seiwa Shoten Publishers.16

Validation

Agreement with SCID-P

We investigated the validity of the MINI Japanese version16 in relation to SCID-P3 as the ‘gold standard’, because the agreement of the English version of the MINI with SCID-P was investigated.12,13 Then we compared the agreement values between the English and Japanese versions.

To ensure an adequate number of samples for each of the major psychiatric disorders, we designed the study to include at least 30 cases of major depressive disorder, 30 cases of anxiety disorder, 20 cases of psychotic disorder, and 10 cases of alcohol or drug dependence disorder. To meet these sampling criteria, inpatients were recruited on a consecutive basis by independent clinicians at Showa University Hospital. All subjects were 18 years old or older. Those with dementia, mental retardation, language problems or serious medical illnesses were excluded.

Each patient was administered the MINI and SCID-P by one interviewer. The patients were randomized regarding which instrument was administered first.

Agreement with expert's opinion

To ensure an adequate number of samples for each of the major psychiatric disorders, we designed the study to include at least 30 cases of major depressive disorder, 30 cases of anxiety disorder and 20 cases of psychotic disorder. To meet these sampling criteria, new outpatients were recruited on a consecutive basis by independent clinicians at Showa University Hospital. All subjects were 18 years old or older. Those with dementia, mental retardation, language problems or serious medical illnesses were excluded.

This study was designed to examine whether MINI-based diagnoses by non-expert psychiatrists were compatible with diagnoses by expert psychiatrists. The diagnostic standard for this study was based on an expert's opinion. All the expert psychiatrists had 15 or more years of experience in their field (i.e. one professor, three associate professors and two assistant professors of psychiatry). The experts who were blind to the results of MINI provided DSM-IV-based diagnosis using whatever source of information they considered to be the most appropriate and were accustomed to, such as interviews with patients, interviews with patients and families, open questions or diagnostic instruments. We did not assess the interrater reliability of DSM-IV-based diagnoses made by experts.

Reliability

Seventy-seven patients were recruited. All the patients were inpatients of Showa University Hospital and had to be 18 years old or older (duration of hospitalization was 42.2 ± 27.4 days). Those with dementia, mental retardation, or serious medical illnesses were excluded.

Interrater reliability was examined by comparing each diagnosis made by the first rater with each diagnosis made by the second rater. The first rater questioned a patient according to the MINI, and subsequently, both the first and second raters rated the patient. Test–retest reliability was examined by comparing each diagnosis made by the first rater with each diagnosis made 1–2 days later by a third rater blind to the results of initial interview. This method of examining the stability of the instrument was more stringent than the conventional procedure in that different examiners were used to conduct the test and retest interviews.

Interviewers and training

Fifteen non-expert psychiatrists who had less than 2 years experience participated in data collection. To ensure comparability between raters, all the raters received a training packet for MINI and a 1-day training session in the use of SCID-P.

Statistical analysis

All diagnoses of patients were analyzed. The validity of diagnosis by the MINI was tested by comparing it with its corresponding diagnoses by SCID-P and by an expert. Diagnostic concordance was assessed using unweighted kappa values, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV).

Diagnostic concordance was calculated using Cohen's kappa value,17,18 which limits chance agreement. In modest-sized samples when a diagnosis occurs at a very low base rate, kappa variability is high.19,20 In general, it was recommended that kappa be calculated only for diagnoses with a frequency of 5% or more. The following conventions were used: kappa values greater than 0.75 were interpreted as indicating excellent agreement, kappa values of 0.45 or lower, poor agreement, and intermediate values, fair to good agreement.20

Sensitivity is the probability of correctly diagnosing a patient who has a disorder (using the SCID-P diagnosis and the expert's opinion as the criterion). Specificity is the probability of correctly diagnosing that a subject does not have a disorder. PPV is the probability that a subject who is identified as having a disorder by MINI actually has a disorder. NPV is the probability that a subject who is identified as not having a disorder actually does not have a disorder.

When the rate of agreement between the diagnosis by SCID-P and the expert's opinion as the criterion for each of these diagnoses was <5%, kappa values were interpreted with caution.

For the interrater and test–retest reliabilities, we used the kappa coefficient.

Ethics

Participation in the study was voluntary. Each subject provided written informed consent after receiving a complete description of the study. The protocol was approved by the ethics committee of Showa University Hospital.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSIONS
  8. ACKNOWLEDGMENTS
  9. REFERENCES

Validation

Agreement with SCID-P

Eighty-two subjects (30 males, 52 females) completed this study. Ages ranged from 18 to 74 years with a mean age of 40.9 (SD = 15.2) years.

Three of the 16 MINI diagnoses (post-traumatic stress disorder, alcohol abuse and drug abuse) were excluded from the validity analysis, because no patient met SCID-P criteria for these three diagnoses.

The results of the comparison of the MINI with the SCID-P are summarized in Table 1. In general, MINI diagnoses were characterized by kappa values indicating good or excellent agreement. Kappa values indicated excellent agreement (more than 0.75) for  major depressive disorder, panic disorder, agoraphobia, obsessive compulsive disorder, psychotic disorder and alcohol dependence. They indicated good agreement (0.60–0.74) for mania and bulimia nervosa. They indicated acceptable agreement (0.45–0.59) for dysthymia, social anxiety disorder, generalized anxiety disorder, drug dependence and anorexia nervosa. Sensitivities were excellent (more than 0.75) for all but three disorders (for dysthymia, drug dependence, and anorexia nervosa). Specificities and NPVs were 0.85 or higher across all the diagnoses; in particular, NPVs were 0.95 or higher. PPVs were excellent (more than 0.75) for major depressive disorder, dysthymia, panic disorder, agoraphobia, psychotic disorder, alcohol dependence and bulimia nervosa. They were good (0.60–0.74) for obsessive-compulsive disorder and anorexia nervosa. They were acceptable (0.45–0.59) for mania, social anxiety disorder, and drug dependence. It was not acceptable (less than 0.44) for generalized anxiety disorder (GAD).

Table 1. Concordance between M.I.N.I. and SCID-P diagnoses Thumbnail image of

Kappa values for the diagnoses of mania, generalized anxiety disorder, drug dependence and bulimia nervosa were interpreted with caution because the rate of meeting SCID-P criteria and the expert's opinion for each of these diagnoses was < 5%.

Agreement with an expert's opinion

A total of 169 (63 males and 106 females; mean age ± SD: 39.6 ± 16.4 years) patients were administered the MINI and subsequently rated by an expert.

Five of the 16 MINI diagnoses (post-traumatic stress disorder, alcohol abuse, drug dependence, drug abuse, and anorexia nervosa) were excluded from the validity analysis, because no patient met the expert's opinion for these five diagnoses.

The results for 11 disorders are summarized in Table 2. In general, kappa values indicated poor agreement for most diagnoses. Agreement was good (0.6–0.74) for psychotic disorder. Kappa values indicated acceptable agreement (0.45–0.59) for panic disorder and bulimia nervosa. Sensitivity was excellent (more than 0.75) for all but three disorders. They were acceptable for dysthymia and bulimia nervosa, and poor for generalized anxiety disorder. Specificity was excellent (0.80 or greater) for all except for major depressive disorder. NPVs were also excellent (0.85 or greater) for all diagnoses. PPVs were good for psychotic disorder, acceptable for panic disorder, social anxiety disorder and bulimia nervosa, and poor for the remaining seven disorders.

Table 2. Concordance between the Mini-International Neuropsychiatric Interview and the expert's diagnoses Thumbnail image of

The diagnoses based on the expert's opinion included 17 adjustment disorders, 10 somatoform disorders, five personality disorders and five other diagnoses.

Kappa values for diagnoses of dysthymia, mania, agoraphobia, obsessive-compulsive disorder, alcohol dependence and bulimia nervosa were interpreted with caution because the rate of meeting the expert's opinion for each of these diagnoses was < 5%.

Reliability

Interrater and test-retest reliabilities

Seventy-seven (29 males and 48 females; mean age ± SD: 54.3 ± 12.6 yeas) patients participated in the reliability study.

Four of 16 MINI diagnoses (post-traumatic stress disorder, drug dependence, drug abuse and alcohol abuse) were excluded from reliability analysis, because none of the patients met MINI criteria for these four diagnoses.

Table 3 shows kappa values for the interrater reliability tests. All of the kappa values indicated excellent agreement (more than 0.75) except those for dysthymia (0.74) and generalized anxiety disorder (0.72).

Table 3. Interrater and test–retest reliabilities of the Mini-International Neuropsychiatric Interview. Thumbnail image of

Table 3 also presents kappa values for comparisons between initial and retest MINIs. Kappa values indicated excellent agreement (more than 0.75) for three of the 12 diagnoses (i.e. major depressive disorder, panic disorder and generalized anxiety disorder), good agreement (0.60–0.74) for six diagnoses, and acceptable agreement (0.45–0.59) for two diagnoses. Only one kappa value (dysthymia) was less than 0.45.

Length of interview

Some sections of the SCID-P (the introductory clinical and demographic sections and the somatoform disorder and adjustment disorder modules) do not have corresponding sections in MINI To facilitate comparison of the administration time, these sections of SCID-P were not administered in this study, artificially shortening the length of SCID-P. Nevertheless, the mean time duration of MINI was about 40% of SCID-P (18.8 ± 6.8 min vs. 45.4 ± 12.8 min).

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSIONS
  8. ACKNOWLEDGMENTS
  9. REFERENCES

Overall, the results were positive. The data suggest that the MINI Japanese version succeed in reliably and validly eliciting symptom criteria used in making DSM-III-R1 diagnoses, and can be performed in less than half the time required to administer the SCID-P. Eighty-two patients were administered both the MINI and the SCID-P. For most diagnoses, concordance between the two diagnostic interviews was good and excellent. Table 1 shows kappa values indicating agreement between the Japanese or English version of MINI and SCID-P.13 The Japanese MINI demonstrated similar, if not higher, kappa values as those of the English version. However, for GAD, drug dependence and anorexia nervosa, the kappa values were lower (about 0.50). The reason for this discrepancy was the small number of patients who were diagnosed positively for these disorders. Higher kappa values may have been obtained for these diagnoses if the base rate of patients with these disorders had been higher.

Small differences between the MINI and the SCID-P were consistently in the direction of the MINI being slightly more overinclusive in the diagnoses, hence the higher rate of false positive results and lower PPV scores for some diagnoses, and in contrast, the lower rate of false negative and perfect NPV scores for almost all diagnoses. When Sheehan et al. started developing the MINI, they decided to design each disorder module to be slightly overinclusive, if necessary, to accept a few more false positives results instead of missing true cases.11–13 They omitted disability, illness, and drug rule out questions from the MINI. Leon et al. suggested that patients with false positive screening results for at least one psychiatric diagnostic criterion had significantly greater functional impairment and higher rates of recent use of mental health services than subjects with true negative results of the screening.21 The MINI indeed demonstrated a lower rate of false negative results, but there were few patients with false negative results. For instance, three patients were false negative for anorexia nervosa (Table 1). These three were diagnosed negative for this disorder by the MINI because screening questions of anorexia nervosa module of the MINI strictly required patient's weight loss to be equal to or below the 85% threshold corresponding to his or her height. Some screening questions must be administered attentively. Despite a reduction of more than 50% in administration time of the MINI compared with that of the SCID-P, the sensitivity and specificity of the MINI were excellent.

Agreement between MINI-based diagnoses and the expert psychiatrist's diagnoses was poor for most diagnoses. However, sensitivity, specificity and NPVs were excellent in contrast to PPVs that were not good. The reason for these discrepancies was the large number of patients who were diagnosed positively by the MINI but negatively by an expert (high rates of false positive). Clinicians often make an intuitive diagnosis without checking all the diagnostic criteria. When they do use DSM-IV, their interpretations of criteria and the qualitative clinical content often depend on their own clinical experience and expertise. Generally, they diagnose only one disorder and they do not diagnose comorbidity.  For  instance,  when  an  expert  diagnosed  a patient for anorexia nervosa, the expert did not attempt to diagnose the patient for major depressive disorder, even when the patient had major depressive episode. In this study, the interrater reliability of diagnoses made by experts was unclear. There was some possibility that a poor interrater reliability of DSM-based diagnoses made by experts might have an influence on the poor agreement between MINI-based diagnoses and the expert's diagnoses. Sheehan et al. also reported that the agreement of kappa values between MINI-based diagnoses and the expert's opinions was not very good (0.41–0.68).13 However, Loze et al. reported that the agreement between MINI-based diagnoses and the clinician's opinions was high (kappa values ranging between 0.76 and 1.00).22 The validity of such a method indicating the concordance of a structured interview diagnosis and an expert's opinion is debatable.23

The interrater reliability was excellent. The test–retest reliability was inferior to the interrater reliability, but it was sufficiently excellent in this sample. Because a different third interviewer was used for the retest (introducing an additional potential source of error), the analysis would be expected to produce a very conservative estimate of stability of MINI diagnoses.

Some sections of the SCID-P were not administered in this study; nevertheless, the mean duration of the MINI administration was about 40% that of the SCID-P. Because of the pressure to spend less time evaluating patients, clinicians can benefit from a tool that would help them use their time more efficiently and maintain or improve their level of diagnostic accuracy.

The MINI is gaining rapid acceptance internationally, being translated into at least 30 languages, mostly European languages, by 2004. The depression module of the MINI was used to screen for depression in a survey of 78 463 adults in a European community.24 The MINI has been used by many academic researchers for the screening of homogenous samples in clinical trials, comorbidity studies and epidemiologic studies (i.e. in general population, workplace, school, and primary care settings).25–29

Limitations

There are some shortcomings in the present study. First, the subjects of this study were only psychiatric patients. We need a study that includes non-psychiatric patients (i.e. in a primary care setting) and normal control subjects. Second, the number of subjects in the study of concordance between the MINI and the SCID-P was relatively small. Moreover, there were only a small number of cases for some diagnoses (mania, GAD, drug dependence, anorexia nervosa and bulimia nervosa). Higher kappa values and PPVs may have been obtained for these diagnoses if the base rate of cases had been higher. Third, we need a study that investigates the validity of the MINI in relation to SCID-I30 or CIDI, which corresponds to DSM-IV.

CONCLUSIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSIONS
  8. ACKNOWLEDGMENTS
  9. REFERENCES

Overall, the data suggest that the Japanese version of the MINI succeeds in reliably and validly eliciting symptom criteria used in making DSM-III-R1 diagnoses, and can be performed in less than half the time required for SCID-P. Although the MINI provides less disorder subtyping than SCID-P, it covers a much broader range of diagnoses than other short structured interviews such as PRIME-MD. The MINI or some modules of the MINI can be used by academic researchers and pharmaceutical companies for a rapid screening of homogenous samples for clinical trials and epidemiologic studies. The MINI has potential applications as a diagnostic screening tool for psychiatric disorders in primary care settings.

ACKNOWLEDGMENTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSIONS
  8. ACKNOWLEDGMENTS
  9. REFERENCES

This research was supported in part by grants from the Glaxo Smith Kline Pharmaceutical. We are grateful to Professor David V. Sheehan, University of South Florida  College  of  Medicine,  Tampa,  Florida,  USA,  and  to our colleagues, Drs Hideki Akiniwa, Masatoshi Hayashi, Taro Ikawa, Takahiro Kato, Rie Kobayashi, Yuki Kudo, Yuka Kuwakado, Kentaro Matsumaru, Tomoko Nagai, Toshimi Owashi, Osamu Takashio, Rieko Sakamoto, Taro Takahashi and Hiroki Yamada.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSIONS
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  • 1
    American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 3rd edn, Revised. American Psychiatric Association, Washington, DC, 1987.
  • 2
    Spitzer RL, Williams JB, Gibbon M, First MB. The structured clinical interview for DSM-III-R (SCID). I. History, rationale, and depression. Arch. Gen. Psychiatry 1992; 49: 624629.
  • 3
    Spitzer RL, Williams JBW, Gibbon M, First MB. Structured Clinical Interview for DSM-III-R (SCID) Patient Edition with Psychotic Screen. American Psychiatric Press, Washington, DC, 1990. Japanese language edition: Takahashi T, Hanada K, Ono Y (eds). Igaku-Shoin, Tokyo, 1992 (in Japanese).
  • 4
    World Health Organization, WHO. The Composite International Diagnostic Interview (CIDI), Authorized Core, Version 1.0. WHO, Geneva, 1990.
  • 5
    Robins LN, Wing J, Wittchen HU et al. The composite international diagnostic interview. Arch. Gen. Psychiatry 1988; 45: 10691077.
  • 6
    World Health Organization, WHO. Schedules for Clinical Assessment in Neuropsychiatry (SCAN). WHO, Geneva, 1992.
  • 7
    Endicott J, Spitzer RL. A diagnostic interview: the schedule for affective disorders and schizophrenia. Arch. Gen. Psychiatry 1978; 35: 837844.
  • 8
    Broadhead WE, Leon AC, Weissman MM et al. Development and validation of the SDDS-PC screen for multiple mental disorders in primary care. Arch. Fam. Med. 1995; 4: 211219.
  • 9
    Weissman MM, Olfson M, Leon AC et al. Brief diagnostic Interviews (SDDS-PC) for multiple mental disorders in primary care. Arch. Fam. Med. 1995; 4: 220227.
  • 10
    Spitzer RL, Williams JB, Kroenke K et al. Utility of a new procedure for diagnosing mental disorders in primary care. The PRIME-MD 1000 study. JAMA 1994; 14: 17491756.
  • 11
    Lecrubier Y, Sheehan DV, Weiller E et al. The Mini International Neuropsychiatric Interview (MINI). A short diagnostic structured interview: reliability and validity according to the CIDI. Eur. Psychiatry 1997; 12: 224231.
  • 12
    Sheehan DV, Lecrubier Y, Harnett Sheehan K et al. The validity of the Mini International Neuropsychiatric Interview (MINI) according to the SCID-P and its reliability. Eur. Psychiatry 1997; 12: 232241.
  • 13
    Sheehan DV, Lecrubier Y, Sheehan KH et al. The Mini International Neuropsychiatric Interview (M.I.N.I.): the development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. J. Clin. Psychiatry 1998; 59: 2233.
  • 14
    American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 4th edn. American Psychiatric Association, Washington, DC, 1994.
  • 15
    Brislin RW. Back-translation for cross-cultural research. J. Cross-Cult. Psychol. 1986; 1: 185216.
  • 16
     Otsubo T, Miyaoka H, Kamijima K (eds). M.I.N.I. Brief Diagnostic Structured Interview for Mental Disease. Seiwa Shoten Publishers, Tokyo, 2000 (in Japanese).
  • 17
    Cohen J. A coefficient of nominal agreement for nominal scales. Psychol. Meas. 1960; 20: 3746.
  • 18
    Cicchetti DV, Feinstein AR. High agreement but low kappa. II. Resolving the paradoxes. J. Clin. Epidemiol. 1990; 43: 551558.
  • 19
    Fleiss JL. Statistical Methods for Rates and Proportions, 2nd edn. John Wiley & Sons, New York, 1981.
  • 20
    Shrout PE, Spitzer RL, Fleiss JL. Quantification of agreement in psychiatric diagnosis revised. Arch. Gen. Psychiatry 1987; 44: 172177.
  • 21
    Leon AC, Portera L, Olfson M et al. False positive results: a challenge for psychiatric screening in primary care. Am. J. Psychiatry 1997; 154: 14621464.
  • 22
    Loze JY, Falissard B, Limosin F, Recasens C, Horreard AS, Rouillon F. Validation of a new diagnostic procedure for DSM IV axis I disorders. Int. J. Methods Psychiatr. Res. 2002; 11: 134141.
  • 23
    Robins LN. The validity of psychiatric diagnosis. In: RobinsLN, BarrettJ (eds). American Psychopharmacological Association Series. Raven Press, New York, 1989; 263278.
  • 24
    Lépine J-P, Gastpar M, Mendelewicz J, Tylee A. Depression in the community: the first pan-European study DEPRES (Depression research in European society). Int. Clin. Psychopharmacol. 1997; 12: 1929.
  • 25
    Balazs J, Lecrubier Y, Crsiszer N, Kosztak J, Bitter IL. Prevalence and comorbidity of affective disorders in persons making suicide attempts in Hungary: importance of the first depressive episode and of bipolar II diagnoses. J. Affect. Disord. 2003; 76: 113119.
  • 26
    Dubini A, Mannheimer, R, Pancheri P. Depression in the community: results of the first Italian survey. Int. Clin. Psychopharmacol. 2001; 16: 4953.
  • 27
    Juang KD, Wang SJ, Fuh JL, Lu SR, Su TP. Comorbidity of depressive and anxiety disorders in chronic daily headache and its subtypes. Headache 2000; 40: 818823.
  • 28
    Nascimento I, Nardi AE, Valence AM et al. Psychiatric disorders in asthmatic outpatients. Psychiatry Res. 2002; 15: 7380.
  • 29
    Olfson M, Lewis-Fernandez R, Weissman MM et al. Psychotic symptoms in an urban general medicine practice. Am. J. Psychiatry 2002; 158: 14121419.
  • 30
     First B, Spitzer RL, Gibbon M, Williams JBW. Structured Clinical Interview for DSM-IV-TR Axis I Disorders, Research Version, Patient Edition (SCID-I/P). Biometrics Research, New York State Psychiatric Institute, New York, 2002.