Is major depression adequately diagnosed and treated by general
practitioners? Results from an epidemiological study
Anna Fernández, M.P.H.a,b,c,⁎, Alejandra Pinto-Meza, Ph.D.a,b,
Juan Angel Bellón, Ph.D., M.D.b,d, Pere Roura-Poch, M.P.H., M.D.e,
Josep M. Haro, Ph.D., M.D.a,f, Jaume Autonell, M.D.a,b, Diego José Palao, Ph.D., M.D.g,
María Teresa Peñarrubia, M.D.h, Rita Fernández, M.D.h, Elena Blanco, M.D.h,
Juan Vicente Luciano, Ph.D.a,b, Antoni Serrano-Blanco, Ph.D., M.D.a,b
aSant Joan de Déu-SSM, Fundació Sant Joan de Déu, Barcelona, Spain
bRed de Investigación en Actividades Preventivas y Promoción de la Salud en Atención Primaria (RedIAPP, Instituto de Salud Carlos III), Spain
cPhD programme, Public Health and Biomedical Research, of the Universitat Autonoma de Barcelona, Spain
dDepartamento de Medicina Preventiva y Salud Pública, Centro de Salud El Palo, Unidad de Investigación del Distrito de Atención Primaria de Málaga,
Universidad de Málaga, Spain
eClinical Epidemiology and Research Department Vic General Hospital, Vic (Barcelona), Spain
fCIBER Salud Mental (Instituto de Salud Carlos III), Spain
gCentre de Salut Mental, Corporació Sanitaria Parc Taulí, Institut Universitari Fundació Parc Taulí-Universitat Autónoma de Barcelona
(Sabadell, Spain), Spain
hCentro de Salud Bartomeu Fabrés Anglada, Gavà, Barcelona, Spain
Received 7 September 2009; accepted 19 November 2009
Objectives: The aim of this study was to (1) to explore the validity of the depression diagnosis made by the general practitioner (GP) and
factors associated with it, (2) to estimate rates of treatment adequacy for depression and factors associated with it and (3) to study how rates of
treatment adequacy vary when using different assessment methods and criteria.
Methods: Epidemiological survey carried out in 77 primary care centres representative of Catalonia. A total of 3815 patients were assessed.
Results: GPs identified 69 out of the 339 individuals who were diagnosed with a major depressive episode according to the Structured
Clinical Interview for DSM-IV Axis I Disorders (SCID-I) (sensitivity 0.22; kappa value: 0.16). The presence of emotional problems as the
patients' primary complaint was associated with an increased probability of recognition. Rates of adequacy differed according to criteria: in
the cases detected with the SCID-I interview, adequacy was 39.35% when using only patient self-reported data and 54.91% when taking into
account data from the clinical chart. Rates of adequacy were higher when assessing adequacy among those considered depressed by the GP.
Conclusion: GPs adequately treat most of those whom they consider to be depressed. However, they fail to recognise depressed patients
when compared to a psychiatric gold standard. Rates of treatment adequacy varied widely depending on the method used to assess them.
© 2010 Elsevier Inc. All rights reserved.
Keywords: Diagnosis validity; Treatment adequacy; Depression; Clinical guidelines; Primary care
Major depression is the most common mental disorder
and one of the most incapacitating . In fact, it has been
suggested that unipolar major depression will be among the
three leading causes of disease burden by 2030 . Twelve-
month prevalence figures in western societies range between
Available online at www.sciencedirect.com
General Hospital Psychiatry 32 (2010) 201–209
⁎Corresponding author. Dr. Antoni Pujadas, 42, 08830 Sant Boi de
Llobregat, Barcelona, Spain. Tel.: +34 93 640 63 50x2543; fax: +34 93 630
E-mail address: email@example.com (A. Fernández).
0163-8343/$ – see front matter © 2010 Elsevier Inc. All rights reserved.
3.6% and 9.1%, and have been estimated at 4% in Spain
[3–5]. Of those, 60% do not receive any kind of health care
for this condition . On the other hand, most of those who
have expressed their need for care will only be treated by a
general practitioner (GP) [7,8]. In recent years many studies
have focused their attention on GPs' ability to recognize
major depression. These studies systematically show that
less than half of the depressed patients who attend primary
care (PC) are recognized and diagnosed [7–13]. However,
the implications of this low recognition are not clear.
Different studies have pointed out that unrecognised cases
tend to be less severe, so detection would be simply a marker
of depression severity [10,14,15], as detected and undetected
depression have a similar course [14,16,17]. The fact that
recognition of depression does not impact its course could be
related to the kind of treatment that recognized cases receive.
Literature focused on the study of treatment adequacy for
depression systematically points out that a large part of
patients with depression receive treatment that does not meet
the criteria recommended by clinical guidelines. Rates of
treatment adequacy for depression in the PC setting vary
widely between 8% and 75% [18–35]. These differences
could be explained by examining the definition used, but
also by the treatment adequacy documentation method:
informed by the patient or taken from the clinical chart.
Adherence to clinical guidelines is important because, in
theory, this guarantees the best available therapeutic option.
In the case of depression, it seems to be associated with
improved outcomes in PC patients with depression at 12, 18
and 24 months . Moreover, Andrews et al.  have
shown that if all depressive patients received treatment
according to the clinical practice guidelines, the burden
averted measured in years lived with disability would climb
Although there is a lot of literature focused on adequate
management of depression, a vast part of these studies has
been carried out in the United States. Few is known about
this topic in countries with a National Health System, in
which GPs play a key role on depression management. On
the other hand, previous studies considered data collected in
the 1990s or early 2000s, that could be considered old.
Moreover, and to the best of our knowledge, no study has
compared how source of data could affect rates of treatment
adequacy (i.e., patient self-reported data vs. data extracted
from the clinical chart). This paper could be among the firsts
assessing this with, recently collected data from the
“Diagnostic and Assessment Study of Mental Disorders in
Primary Care” (DASMAP) project , an epidemiological
study carried out in PC centres in Catalonia.
The study was a face-to-face, cross-sectional survey of a
representative sample of adult patients (18 years or older) at
primary health care centres in Catalonia (Spain). Catalonia is
one of the 17 autonomous regions that comprise Spain, with
a population of 7,134,697 inhabitants according to the 2006
census. In 1981, Spain began a process of decentralization of
health service management and this is now carried out by
each autonomous region. In Spain, health services are funded
at the state level, with nearly universal healthcare free of
charge at the point of use. Data were collected between
October 2005 and March 2006 using a paper-and-pencil
interview. After an appointment with a GP, patients were
invited to participate in the DASMAP study and were
evaluated at the PC centers after written informed consent
had been obtained. Participants and GPs were informed that,
just for reasons of patient safety (suicidal ideation),
information gathered in the interview would not be disclosed
to the GPs. During the study, no information regarding any
patient was disclosed to the GPs. The study was approved by
Sant Joan de Déu Foundation Ethics Board.
A stratified multistage probability sample without
replacement was drawnin the DASMAP study.Replacement
was prohibited to ensure that every individual had a known
probability of selection. The sampling frame was the seven
health regions of Catalonia. Stage 1: selection of the PC
centres within each health region (there were 350 PC centres
in Catalonia in 2005). The number of PC centres to be
selected in each region was proportional to the population of
the region. However, in order to have a minimum set of
interviews from even the smallest regions, at least 6 PC
centres were chosen per region. The probability of selection
of each PC centre was related to the population of the
catchment area covered by the centre. Eighty PC centres
were selected for participation and two refused (97.5%
acceptance rate). Stage 2: all GPs (n=913) at the PC centres
were invited to participate. A total of 618 GPs (67.7%) took
part in the study. Stage 3: patients were selected with a
systematic sampling strategy from the daily list of all patients
with an appointment with each of the participating GPs.
Inclusion criteria were: being older than 18 years old and
having made at least one visit with the GP. Only patients
with severe cognitive impairment were excluded. A total of
5402 patients were randomly selected. Of those, 654 (12.1%)
did not keep their appointment with the GP, so they were not
invited to join the DASMAP study. A total of 4748 were
invited to participate. A further 764 (16.1%) declined to
participate, and 164 (3.5%) were excluded because they
showed cognitive impairment severe enough to preclude an
adequate interview, leaving a study sample of 3820
participants from 78 PC centres. One of the PC centres
was excluded from the statistical analyses because of data
loss. The final analysed sample consisted of 3815 patients
(80.3% of the patients initially invited) with a mean age of
54.3 years (S.D.=17.31; range: 18–97 years). Females
represented 63% of the sample. A total of 64.4% were
married or living with someone, and nearly 49% had primary
202A. Fernández et al. / General Hospital Psychiatry 32 (2010) 201–209
studies. Thirty-four percent of the sample had a paid
employment at the moment of the interview; 11.6% had a
paid employment but were in sick leave, being the other
participants in other situations (1.5% were students, 13.4%
homemakers, 4.6% were unemployed, 29.7 % were retired,
3.4% had a permanent disability and just in 1.3%, data were
missing. More information about sociodemographics could
be obtained on Serrano et al. .
All assessments were performed face-to-face by a group
of 20 trained clinical psychologists. All of them received a
2-day training course provided by the study investigators
(A.S.B., D.P., A.P.M.).
Major depression and dysthymic disorder diagnoses
were assessed with the Spanish version of the Structured
Clinical Interview for DSM-IV Axis I Disorders (SCID-I)
. This instrument allows diagnoses according to DSM-IV
Chronic physical conditions were assessed using a
checklist with questions about a wide range of conditions
including asthma, bronchitis, ulcer, constipation, high blood
pressure, heart disease, heart attack, migraines, stroke,
allergies, arthritis, back pain, cataracts, diabetes and neck
pain. Participants were categorized, by means of self-report,
according to whether they had each chronic physical
condition or not.
All respondents were also asked about their main
complaint: physical, emotional, administrative, or a combi-
nation. In addition, they were asked whether they had visited,
in the 12 months prior to the survey, a mental health
professional (psychiatrist or psychologist).
Self-perceived disability was assessed by means of three
items of the Sheehan Disability Scale. The questions
evaluated interference in three domains: work, social life,
and family life. Participants could answer using a scale
ranging from 0 to 10. The total score is calculated by the sum
of the scores in the three items, ranging from 0 to 30, with
higher figures indicating greater disability [41,42].
Diagnoses made by the GP were extracted from the
clinical chart and recoded into a dichotomous variable that
included the presence or absence of a depression diagnosis.
GP diagnoses were based on the International Classifica-
tion of Diseases (ICD), in its ninth or tenth version, or on
The International Classification of Primary Care second
version (ICPC-2) by the World Organization of National
Colleges, Academies and Academic Associations of
General Practitioners/Family Physicians. Table 1 shows
the different codes included.
Guideline-concordant treatment was defined as receiv-
ing a recommended antidepressant at minimum daily
dosage plus at least 4 visits to the prescribing GP during
the past 12 months or, in absence of medication, at least
eight visits to the GP [43,44]. Two different scenarios
were created for diagnoses: (1) major depression accord-
Prevalence, sensitivity and number missed to screen in order to miss one case of the 12 months depressive disorders
% of cases missed
by the GP
Missed cases as %
of total screen
Number needed to
screen to miss 1 case
ICD-9: 296.2, 296.20, 296.21,
296.22, 296.23, 296.25, 296.25, 296.26, 296.3, 296.31,296.32, 296.34, 296.35, 296.36, 300.4, 309.1, 311
ICD-10:F.32, F32.0, F32.1,
F32.2, F32.8, F32.9, F33,F33.0, F33.1, F33.2, F33.4, F33.8, F33.9, F43.8, F34.9, F38, F38.0, F38.1, F38.8,F38.9ICPC-2: P76
disorder or other depressive
ICD-10: F34.0, F34.1, F41.2
All the codes above mentioned
1Cases registered in the clinical chart by the GP.
203A. Fernández et al. / General Hospital Psychiatry 32 (2010) 201–209
ing to a psychiatric gold-standard (the SCID-I) and (2)
major depression according to the GP. Moreover, in order
to evaluate how a different method could influence
results, treatment adequacy was assessed by means of
two methods for each scenario: (a) using self-reported
data (that is, using the number of sessions, antidepres-
sants and dosage that interviewees reported they were
prescribed), and (b) using data extracted from the clinical
chart (filled out by GPs).
2.3. Statistical analyses
The sensitivity, specificity, and positive and negative
predictive value associated with physician identification of
depression were calculated by standard methods. Addi-
tionally, we reported another indicator of the acceptability
of practice, which set in the denominator the total
participants, in our case, 3815, rather than the number of
cases. As such, we were also able to calculate the number
of patients who would need to be screened in order to
diagnosis one case of depression that would otherwise be
missed . There were 695 of the 3815 participants
(16.45%) who did not have GP diagnosis information
(information missing from the clinical chart). In order to
minimize their effect, we excluded these cases when
carrying out sensitivity data analysis.
In order to ascertain factors associated with (a) major
depression diagnosis made by the GP and (b) a concordant
SCID-I/GP major depression, we used multilevel logistic
regression with GP and PC centre as random factors. Firstly,
we tested whether multilevel logistic regression was more
appropriate than standard logistic regressions. When we
compared the null models, including the PC centre as a
random factor, likelihood ratio tests were very significant for
all models. Secondly, we compared null models that
included PC centre and GP as random factors versus those
with only PC centre, and likelihood ratio tests were not
significant. That is, GP did not add additional information, so
it was excluded as a random factor.
Multilevel logistic regression models, considering PC
centres as random effects, were also tested to study the
effects of clinical and sociodemographic variables on the
likelihood of receiving guideline-concordant care. Never-
theless, null models were not statistical significant, so we
ruled out these analyses, using standard logistic regressions.
The statistical analyses were conducted using STATA 10.1.
All significance tests were done using two-sided tests
evaluated at 0.05 significance level.
3.1. Recognition rates of depression
Twelve-month prevalence of major depression according
to the SCID-I was 9.60% (95% CI 8.00–11.19%; n=339).
GP identified 69 out of the 339 individuals diagnosed with
the SCID-I (sensitivity 0.20 or 0.22 when excluding missing
diagnoses in the clinical chart). As expected, due to the low
prevalence of depression, the specificity was high: 0.94,
whereas the positive predictive value was 0.25. The kappa
value was 0.16.
Among all consultations, 7.08% (270 out of 3815) failed
to identify a case of major depression according to the
SCID-I. This means that one patient with major depression
was undetected or unregistered in the clinical chart every
Sensitivity rate increased to 0.36 (0.40 when excluding
missing data) when considering a pooled category that
included dysthymic disorder, major depression or other
depressive symptoms (Table 2).
3.2. Factors associated with recognition rates
3.2.1. Who did the GP identify as having a diagnosis of
When comparing those identified by the GP as depressed
with those not identified (independently of diagnosis made
Factors associated to the diagnosis of depression made by the general
practitioner and to an adequate diagnosis according to the SCID-I
Variable Model 1:
according to the
OR 95% CIOR 95% CI
Chronic physical illness
Psychiatric family history
Main complaint: mental or emotional problems
Variance explained by PC center 1.22 (0.73–2.03)
3.83 2.70–5.43⁎⁎⁎ 2.90 1.39–6.02⁎⁎
204A. Fernández et al. / General Hospital Psychiatry 32 (2010) 201–209
with the SCID-I), data showed that having past depressive
episodes (OR=2.71, 95% CI 1.99–3.70), reporting mental or
emotional problems as the main complaint (OR=3.83, 95%
CI 2.70–5.43), and being more disabled according to the
Sheehan scale (OR=1.04, 95% CI 1.03–1.06) were associ-
ated with an increased probability of being diagnosed as
depressed (Model 1 in Table 2).
3.2.2. Who did the GP adequately identify as depressed,
according to the SCID-I?
Among those with major depression according to the
SCID-I, only the presence of mental or emotional problems
as the patients' main complaint was associated with an
increased probability of adequate recognition (OR=2.90,
95% CI 1.39–6.24) (Model 2 in Table 2).
3.3. Proportion of minimally adequate treatment among
those with depression according to the SCID-I: Scenario 1
Among those with major depression according to the
SCID-I (n=339), rates of treatment adequacy oscillated
between 39.35% (using self-reported data) and 54.91%
(using data extracted from the clinical chart). Concordance
between these methods was 62.96%, with a kappa value
3.4. Factors associated with minimally adequate treatment
3.4.1. Using self-report data
Model 1 in Table 3 shows factors related to treatment
adequacy among those with major depression according to
the SCID-I using self-report data. Factors associated to
adequacy were self-perceived disability (OR=1.04, 95% CI
1.01-1.07) and reporting use of outpatient mental health
services (OR=2.95, 95% CI 1.88–4.62).
3.4.2. Using data from the clinical chart
On the other hand, in the Model using data from the
clinical chart, having a SCID-I/GP major depression
concordant diagnosis (OR=1.80, 95% CI 1.08–3.01) and
self-perceived disability (OR=1.03, 95% CI 1.00–1.05) were
Factors associated to treatment adequacy among those depressed according to the SCID-I and among those identified as depressed by the general practitioner
Scenario 1: depression according
to the SCID-I (n=339)
Scenario 2: depression according to the
general practitioner (n=270)
1. Self-informed data2. Clinical chart 3. Self-informed data4. Clinical chart
treatment (95% CI)
Chronic physical illness
Psychiatric family history
Concordant major depression diagnosis?
Use of outpatient mental health services?
39.35% (33.46–45.24%)54.91% (49.48–60.36%) 41.76% (35.77–47.99%) 63.36% (57.23–69.08%)
OR95% CI OR95% CI OR 95% CI OR95% CI
1.61 0.88–3.04 0.76–2.150.54–2.470.77–3.41
1.01 0.71–1.940.61–1.41 0.57–1.70 0.56–1.82
205A. Fernández et al. / General Hospital Psychiatry 32 (2010) 201–209
associated with increased probability of receiving adequate
treatment (Model 2 on Table 3).
3.5. Proportion of minimally adequate treatment among
those with depression according to the GP: Scenario 2
A second scenario was created to analyse the proportion of
minimally adequate treatment among those considered
depressed by the GP (n=270). Rates of treatment adequacy
(using data from the clinical chart). Agreement between these
methods was 58.72%, with a kappa value of 0.24.
3.6. Factors associated with minimally adequate treatment
3.6.1. Using self-report data
Model 3 on Table 3 deals with participants identified
as depressed by GPs. When using self-report data, self-
perceived disability (OR=1.08, 95% CI=1.04–1.11) and
reported chronic physical illness (OR=3.21 95%
CI=1.00–10.26) increased the likelihood of receiving
3.6.2. Using data from the clinical chart
No clinical or sociodemographic factors were associated
with treatment adequacy when using data extracted from the
clinical chart. The only factor that was associated with an
increased probability of receiving guideline-concordant
treatment was having made at least one visit to a mental
health professional (OR=2.11, 95% CI 1.05–4.23) (Model 4
on Table 3).
4.1. The recognition of major depression by the GP
Our findings suggest that GPs in Catalonia still poorly
clinical guidelines. Adequate recognition of patients with
depression by the GP may be as low as 22% and only a
quarter of the cases diagnosed as depressed by physicians
Thompson et al.  found in the United Kingdom. In their
as gold standard a score of N8 on the subscale of Depression
of the “Hospital Anxiety and Depression” scale by Zigmond
and Snaith) was nearly 12, very similar to the 14 people we
needed to screen in our study to miss one case. Nevertheless,
when looking at our data in detail, participants diagnosed as
depressed by their GP were more likely to report past history
of depressive episodes and also showed more disability
according to the Sheehan scale. Moreover, they diagnosed as
depressed those participants whose main complaint was
related to mental or emotional problems. This is quite
interesting ifweconsiderthat,when using theSCID-Iasgold
standard, an adequate diagnosis of depression was also more
likely to be made in patients presenting mental or emotional
problems as their primary complaint. This has been indicated
public about the core symptoms of depression and fighting
useful tool for increasing recognition. Nevertheless, GPs also
need to be trained in “listening” to patients' emotional
complaints, as one of the barriers to accurate recognition and
treatment is related to GPs' attitudes toward depression
It is important to note here that diagnoses for research
purposes may constitute a very different activity than
diagnoses in a clinical setting. In this sense, the SCID-I
could identify as depressed people that are not disabled
enough by this condition to seek treatment, but who are
going to the GP for other health matters. Furthermore, we
know that the SCID-I is not a perfect instrument and that its
reliability is far from ideal [48,49]. Additionally, one
possible explanation for the low concordance between the
SCID-I and GPs diagnoses could be the fact that we are
comparing 4 different classification systems. That is to say,
we are comparing the SCID-I (a psychiatric instrument that
allows a DSM-IV diagnosis) with ICD classifications (in its
ninth or tenth version) or the ICPC that use neither exactly
the same criteria nor the same codification, making
comparisons difficult. On the other hand, it has been
suggested [10,50] that when diagnosing, GPs use a
dimensional point of view instead of a categorical one (as
these classifications systems do). Moreover, it has been
argued that recognition of depression by GP is a process
instead of a one-visit decision, and that many unrecognised
cases are recognized in subsequent visits. If this is the case, it
is possible that some of our undetected cases would be
diagnosed in the future. Nevertheless, Van Os et al. 
showed that after 1 year of follow-up, unrecognized patients
decreased only slightly (from 27% to 20%).
It is important to mention that we are considering the
psychiatric point of view as the gold standard for GPs'
activity. As others have pointed out [51,52], this may be
inadequate, since we are forgetting the context in which GPs
work. Catalan GPs work under extreme pressure (between 7
and 10 min per visit), and this brief period precludes an
adequate psychiatric interview. The time barrier has been
systematically identified in the literature as a serious obstacle
to an adequate diagnosis [53–55]. Another explanation is
that GPs may simply forget to register it on the clinical chart.
A recent meta-analysis  has shown that the method of
documentation of depression recognition by GP has a
significant effect on sensitivity, being higher in those studies
that directly asked GPs about this possible diagnosis than in
those studies that sought the diagnosis in the clinical chart, as
we did. On the other hand, we should bear in mind that
mental disorders are still stigmatised and GPs might hide the
diagnosis from a patient (and may not write it down in the
clinical chart) if they believe that the person may feel afraid
Nevertheless, if we consider that GPs detected nearly
40% of those with depressive symptoms, we could
206A. Fernández et al. / General Hospital Psychiatry 32 (2010) 201–209
hypothesize that they recognized mood disorders, although
they did not discriminate between them. This is very similar
to the results of the above-mentioned meta-analysis on
recognition of depression by nonpsychiatrist physicians that
indicated a sensitivity of 36.4% .
It is also worth pointing out the high proportion of cases
that were considered to be depression by the GP, but not by
the SCID-I (false positives). One qualitative study carried
out in Canada, aiming to explore PC physicians' experiences
of diagnosing and treating depression, suggested that the
main barrier to diagnosis and adequate treatment of
depression was the conflict that arises between the medica-
lized understanding of depression and recognition of the
social context of depression. The results of this conflict are
dissonant descriptions of depression that could complicate
adequate diagnosis as physicians are not able to discriminate
between life problems and medical problems . This
aspect, the medicalization of normal sadness, has also been
for major depression [57,58]. Moreover, GPs could use
depression diagnosis (knowing that patient may not fill
criteria) as a social tool, for example, to allow patients to take
sick leave or because it is the only way to refer them to
specialised care such as mental health services .
4.2. Treatment adequacy
Clinical guidelines are based on scientific evidence about
the treatment efficacy of specific interventions on specific
diagnoses. Nowadays, most of these studies are clinical trials
that use DSM-IV criteria for recruiting samples. That is why
diagnoses based on psychiatric criteria are so important for
providing adequate treatment. A patient with a “specific
diagnosis” is treated under the assumption that the symptoms
will improve if the disease is effectively treated according to
clinical guidelines . In Scenario 1, rates of adequate
treatment oscillated between 40% and 55%; higher than
previously reported in Spain [6,32,34]. Nevertheless, com-
parison is difficult because definitions of adequacy are
different across studies. It is interesting to note that SCID-I/
GP concordant diagnoses are associated with increased
probability of being adequately treated when considering
disability was also associated with increasing the possibility
of adequate treatment for those with major depression
according to the SCID-I, suggesting that GPs may not
make a specific diagnosis but recognize the suffering of these
patients and treat them. On the other hand, when assessing
treatment adequacy in those with major depression according
to GPs (Scenario 2), rates of adequacy ranged between 42%
and 63%. As might be expected, treatment adequacy was
higher among those considered as depressed by the GP.
associated with increased treatment adequacy. This has been
systematically highlighted in the literature [34,60], suggest-
ing that one way to improve the quality of depression
treatments could be collaborative interventions by GPs and
mental health professionals.
4.3. The influence of the method used
Finally, an important issue arising from our results is that
the method used to assess adequacy dramatically affects
adequacy rates. Self-report data could be biased by recall
problems, but data from the clinical chart could be affected
by the extent to which GPs completed it. Both methods have
limitations, but it is likely that assessment of adequacy using
data from the clinical chart is more accurate since it at least
reflects the real treatment provided by GPs, and it is not
affected by recall bias. This could have an impact on future
research, as studies on adequacy carried out with epidemi-
ological samples could be underestimating the proportion of
These results should be interpreted with the following
limitations in mind. First, since the DASMAP project is a
cross-sectional study, we cannot conclude that factors
associated with adequate diagnoses and treatment of patients
with major depression are causally related. Second, our
definition of minimally adequate treatment is imperfect. We
used a loose definition of treatment inadequacy which may
have underestimated it (i.e., we did not consider which kind
of counselling GPs made when not prescribing an antide-
pressant, neither the interval between visits nor the
interaction among prescribed medications). Moreover, we
did not take into account that a treatment could be also
considered inadequate by overutilisation of resources. Third,
due to the study design, we were not able to test whether
adequate recognition and/or treatment were associated with
better outcomes. Fourth, we did not evaluate some relevant
GP variables such as attitudes toward mental disorders or
education in mental health, or organizational variables such
as mean time per visit or availability of multidisciplinary
teams with mental health professionals, which could be
related with adequate recognition and treatment of major
depression. Fifth, we did not assess if GPs referred patients
with depression to a mental health professional, although as a
proxy, we considered if patients consulted with a mental
health professional. Finally, the study had a relatively small
sample size, so results have to be interpreted with caution.
We thank Luis Salvador for their helpful comments on
drafts. This study was funded by the “Direcció General de
Planificació i Avaluació Sanitària — Departament de Salut
— Generalitat de Catalunya” (Barcelona, Spain). A.F. and J.
V.L. are grateful to the “Ministerio de Sanidad y Consumo,
Instituto de Salud Carlos III” (Red RD06/0018/0017) for a
predoctoral and a postdoctoral contract, respectively.
207 A. Fernández et al. / General Hospital Psychiatry 32 (2010) 201–209
 Paykel ES, Brugha T, Fryers T. Size and burden of depressive
disorders in Europe. Eur Neuropsychopharmacol 2005;5:411–23.
 Mathers CD, Loncar D. Projections of global mortality and burden of
disease from 2002 to 2030. PLoS Med 2006;e442:3.
 Haro JM, Palacín C, Vilagut G, et al. Prevalence of mental disorders
and associated factors: results from the ESEMeD-Spain study. Med
Clin (Barc) 2006;26:445–51.
 Demyttenaere K, Bruffaerts R, Posada-Villa J, et al. Prevalence,
Severity, and Unmet Need for Treatment of Mental Disorders in the
World Health Organization World Mental Health Surveys. JAMA
 Gabilondo A, Rojas-Farreras S, Vilagut G, et al. Epidemiology of
major depressive disorders in southern European country: results from
the ESEMeD-Spain Project. J Affect Disord 2009.
 Fernández A, Haro JM, Codony M, et al. Treatment adequacy of
anxiety and depressive disorders: primary versus specialised care in
Spain. J Affect Disord 2006;96:9–20.
 Lecrubier Y. Widespread underrecognition and undertreatment of
anxiety and mood disorders: results from 3 European Studies. J Clin
 Üstün TB, Sartorius N. Mental Illness in General Health Care.
Chichester: John Wiley; 1995.
 Klinkman MS, Coyne JC, Gallo S, Schwenk TL. False positives, false
negatives, and the validity of the diagnosis of major depression in
primary care. Arch Fam Med 1998;7:451–61.
 Thompson C, Ostler K, Peveler RC, Baker N, Kinmonth AL.
Dimensional perspective on the recognition of depressive symptoms
in primary care. The Hampshire Depression Project 3. Br J Psychiatry
 Wittchen HU, Pittrow D. Prevalence, recognition and management of
depression in primary care in Germany: the Depression 2000 study.
Hum Psychopharmacol 2002;17:S1–S11.
 Cepoiu M, McCusker J, Cole MG, et al. Recognition of depression by
non-psychiatric physicians-a systematic literature review and meta-
analysis. J Gen Intern Med 2008;23:25–36.
 Berardi D, Menchetti M, Cevenini N, et al. Increased Recognition of
Depression in Primary Care. Comparison between Primary-Care
Physician and ICD-10 Diagnosis of Depression. Pyschother Psycho-
 Goldberg D, Privett M, Ustun B, Simon G, Linden M. The effects of
detection and treatment on the outcome of major depression in primary
care: a naturalistic study in 15 cities. Br J Gen Pract 1998;48:1840–4.
 Aragonés E, Piñol JL, Labad A, Folch S, Mèlich N. Detection and
management of depressive disorders in primary care in Spain. Int J
Psychiatry Med 2004;34:331–43.
 Dowrick C. Buchan I. Twelve month outcome of depression in general
practice: does detection or disclosure make a difference? BMJ 1995;
for depression: a meta-analysis. CMAJ 2008;178:997–1003.
 Lin EH, Katon WJ, Simon GE, et al. Achieving guidelines for the
treatment of depression in primary care: is physician education
enough? Med Care 1997;35:831–42.
 Katz SJ, Kessler RC, Lin E, Wells KB. Medication management of
depression in the United States and Ontario. J Gen Intern Med 1998;
 Wang PS, Berglund P, Kessler RC. Recent care of common mental
disorders in the United States: prevalence and conformance with
evidence-based recommendations. J Gen Psychiatry 2000;15:
 Young AS, Klap R, Sherbourne CD, Wells KB. The quality of care for
depressive and anxiety disorders in the United States. Arch Gen
 Wang PS, Demler O, Kessler RC. Adequacy of treatment for serious
mental illness in the United States. Am J Public Health 2002;92:92–8.
 Weilburg JB, O'Leary KM, Meigs JB, Hennen J, Stafford RS.
Evaluation of the adequacy of outpatient antidepressant treatment.
Psych Serv 2003;54:1233–9.
 Dobscha SK, Gerrity MS, Corson K, Bahr A, Cuilwik N. Measuring
adherence to depression treatment guidelines in a VA primary care
clinic. Gen Hosp Psychiatry 2003;25:230–7.
 Kessler RC, Berglund P, Demler O, et al. The epidemiology of major
depressive disorder: results from the National Comorbidity Survey
Replication (NCS-R). JAMA 2003;289:3095–105.
 Simon GE, Fleck M, Lucas R, Bushnell DM, LIDO Group. Prevalence
and predictors of depression treatment in an international primary care
study. Am J Psychiatry 2004;161:1626–34.
 Balestrieri M, Carta MG, Leonetti S, et al. Recognition of depression
and appropriateness of antidepressant treatment in Italian primary care.
Soc Psychiatry Psychiatr Epidemiol 2004;39:171–6.
 Wang PS, Lane M, Olfson M, et al. Twelve-month use of mental
health services in the United States: results from the National
Comorbidity Survey Replication. Arch Gen Psychiatry 2005;62:
 Joo JH, Solano FX, Mulsant BH, Reynolds CF, Lenze EJ. Predictors of
adequacy of depression management in the primary care setting.
Psychiatr Serv 2005;56:1524–8.
 Kniesner TJ, Powers RH, Crogham TW. Provider type and depression
treatment adequacy. Health Policy 2005;72:321–32.
 Van Os TWDP, Van den Brink RHS, Van der Meer K, Ormel J. The
care provided by general practitioners for persistent depression. Eur
 Fernández A, Haro JM, Martinez-Alonso M, et al. Treatment adequacy
for anxiety and depressive disorders in six European countries. Br J
 Sewitch MJ, Blais R, Rahme E, Bexton B, Galarneau S. Receiving
guideline-concordant pharmacotherapy for major depression: impact
on ambulatory and inpatient health service use. Can J Psychiatry 2007;
 Pinto-Meza A, Fernández A, Serrano-Blanco A, Haro JM. Treatment
adequacy of pharmacological antidepressant treatment in Spanish
primary care. A 6 months naturalistic follow-up study. Psychiatr Serv
 Mischoulon D, McColl-Vuolo R, Howarth S, et al. Management of
major depression in the primary care setting. Psychother Psychosom
 Hepner KA, Rowe M, Rost K, et al. The effect of adherence to
practical guidelines on depression outcomes. Ann Intern Med 2007;
 Andrews G, Issakidis C, Sanderson K, Corr J, Lapsley H. Utilising
survey data to inform public policy: comparison of the cost-
effectiveness of treatment of ten mental disorders. Br J Psychiatry
 Serrano-Blanco A, Palao DJ, Luciano JV, et al. Prevalence of mental
disorders in primary care: results from the diagnosis and treatment of
mental disorders in primary care study (DASMAP). Soc Psychiatry
Psychiatr Epidemiol 2009.
 First MB, Gibbon M, Spitzer RL, Williams JBW. User's guide for the
Structured Clinical Interview for DSM-IV Axis I Disorders. Research
Version. New York: Biometrics Research; 1996. p. 1996.
 American Psychiatric Association. Diagnostic and Statistical Manual
of Mental Disorders. 4th ed. Washington, DC: American Psychiatric
 Bobes J, Badia X, Luque A, et al. Validación de las versiones en
español de los cuestionarios Liebowitz Social Anxiety Scale, Social
Anxiety and Distress Scale y Seehan Disability Inventory para la
evaluación de la fobia social. Med Clin (Barc) 1999;112:530–8.
 Sheehan DV, Harnett-Sheehan K, Raj BA. The measurement of
disability. Int Clin Psychopharmacol 1996;11:89–95.
 National Institute for Clinical Excellence. Depression: management of
depression in primary and secondary care. Available at www.nice.org.
208A. Fernández et al. / General Hospital Psychiatry 32 (2010) 201–209
 American Psychiatric Association. Practice guideline for the treatment Download full-text
of patients with major depressive disorder. 2nd ed. Washington, DC:
American Psychiatric Association; 2000.
 Araya R, Lewis GH, Rojas G, Mann AH. “Patient knows best”
detection of common disorders in Santiago, Chile: cross sectional
study. BMJ 2001;322:79–81.
 Dowrick C, Gask L, Perry R, Dixon C, Usherwood T. Do general
practitioners' attitudes towards depression predict their clinical
behaviour? Psychol Med 2000;30:413–9.
 Gask L, Dixon C, May C, Dowrick C. Qualitative study of an
educational intervention for GPs in the assessment and management of
depression. Br J Gen Pract 2005;55:854–9.
 Wlliams JB, Gibbon M, First MB, et al. The Structured Clinical
Interview for DSM-III-R (SCID). II. Multisite test-retest reliability.
Arch Gen Psychiatry 1992;49:630–6.
 Segal DL, Hersen M, Van Hasselt VB. Reliability of the Structured
Clinical Interview for DSM-III-R: an evaluative review. Compr
 Peveler R, Kendick T. Treatment delivery and guidelines in primary
care. Br Med Bull 2001;57:193–206.
 Katerndahl DA, Larme AC, Palmer RF, Amodei N. Reflections onDSM
classification and its utility in primary care: case studies in “mental
disorders”. Prim Care Companion J Clin Psychiatry 2005;7:91–9.
 McLeod D, on behalf of The MaGPIe Research Group. General
practitioner recognition of mental illness in absence of a ‘gold
standard’. Aust N Z J Psychiatry 2004;38:789–94.
 Freeman AC, Sweeney K. Why general practitioners do not implement
evidence: qualitative study. BMJ 2001;323(7321):1100–2.
 Pollock K, Grime J. Patients' perceptions of entitlement to time in
general practice consultations for depression: qualitative study. BMJ
 Mechanic D. Barriers to help-seeking, detection, and adequate
treatment for anxiety and mood disorders: implications for health
care policy. J Clin Psychiatry 2007;68(2):20–6.
 Thomas-MacLean R, Stoppard JM. Physicians' constructions of
depression: inside/outside the boundaries of medicalization. Health
 Parker D. Beyond major depression. Psychol Med 2005;35:467–74.
 Horwitz AV, Wakefield JC. The loss of sadness. How psychiatry
transformed normal sorrow into depressive disorders. New York:
Oxford University Press; 2007.
 Kirk SA, Kutchins H. The selling of DSM: the rhetoric of science in
psychiatry. New York: Aldine de Gruyter; 1992.
 Katon W, Von Korff M, Lin E, et al. Collaborative management to
achieve treatment guidelines. Impact on depression in primary care.
209A. Fernández et al. / General Hospital Psychiatry 32 (2010) 201–209