Introduction: Patients seeking psychotherapy often spend time on waitlist (WL), the effect of which is largely unknown. WL patients may forego alternative non-psychotherapeutic assistance and thus do more poorly than had they not been placed on a WL. The course of symptoms might also be influenced by use of antidepressant medication (ADM), an issue that remains unexplored in the literature. Objective: In a naturalistic setting, WL symptom change before inpatient psychotherapy (mean weeks of waiting = 22.6) was assessed in a sample (N = 313) of chronically depressed patients. Methods: Using the Beck Depression Inventory-II, patients’ symptoms were tracked at assessment, when admitted to treatment (i.e., after WL), at posttreatment and 1-year follow-up. Multilevel growth curve analysis was used to examine waitlist change for the whole sample as well as for ADM users and nonmedicated patients. Results: Symptoms were reduced significantly from assessment to admittance (Cohen’s d = 0.47). Symptoms reduced less for ADM users (d = 0.39) than for nonmedicated patients (d = 0.65). Conclusion: The findings indicate that chronically depressed patients experience a decrease in symptoms during WL, quite likely due to treatment expectations. We discuss whether less symptom improvement for ADM users could be attributed to iatrogenic comorbidity and a higher degree of demoralization in this group.

When patients seek help for their mental health concerns, they often spend time on a waitlist (WL) before starting treatment. The act of seeking treatment could facilitate hope and positive expectations for improvement even if treatment will not be available immediately. Also, WL could lead to increased distress because it might reduce patients’ motivation for seeking and receiving alternative care [1, 2]. Distress during WL could also be influenced by the natural course of the disorder [3]; it might be progressive and naturally deteriorating, or episodic and naturally improving. Definitive conclusions about the effects of WL are unavailable because it is not feasible nor ethical to randomly assign patients to a condition where no help will eventually be available. Prior research on the effects of waiting for psychotherapy has not been conclusive. There is some evidence that patients with depression experience short-term improvement on WL [4]. Others claim that WL may be a nocebo [1] because it results in deterioration relative to other control conditions such as placebos and no treatment [5, 6]. We are not aware of studies showing that being on a WL systematically results in deterioration. In the present study, we explored in a naturalistic context, symptom change while being on WL for patients with chronic depression.

Many depressed patients use antidepressant medication (ADM) and chronically depressed patients often present for psychotherapy on a course of ADM [7, 8]. It is unclear how treatment with ADMs impacts long-term outcomes in depression, and it is difficult to predict how patients using ADM will respond during WL compared to nonmedicated patients. One hypothesis could be that long-term ADM users develop tolerance and/or resistance to ADM and thus lose their beneficial effects [9‒11] but are reluctant to discontinue out of fear that symptoms will be even worse [12]. If this is the case, patients on non-beneficial ADM may be demoralized as unsuccessful treatments violate expectations of improvement and may give patients an indication that their condition is unresponsive to treatment [13]. Thus, they may experience less hope and positive expectations facing new treatment attempts compared to nonmedicated patients.

We investigated the following research questions: (1) How do depressed patients’ symptoms change when they are waiting to enter a psychotherapy program? (2) Does ADM-use affect the course of symptoms while waiting for treatment?

Study Design and Participants

In this naturalistic study, we assessed symptom change during WL, pending admission to a 12-week inpatient treatment program for adult patients with persistent depressive disorder or recurrent major depressive disorder at an inpatient treatment facility in Norway. Exclusion criteria at the depression unit were psychosis, cluster A and B personality disorder, bipolar disorder, ongoing substance abuse, and organic brain disorder. Patients who during treatment had been re-diagnosed with another primary diagnosis than depression, and patients using psychotropic medication not prescribed for antidepressive purposes, were excluded from analyses. Patients using other non-ADM medication for the purpose of treating depression were included in the analyses (i.e., lamotrigine, aripiprazole, quetiapine). Of the final sample (N = 313), 144 (46%) patients were on medication used for antidepressant purposes, while 169 (54%) patients were unmedicated.

Procedures and Outcome

Primary outcome for depressive symptoms was Beck Depression Inventory-II (BDI-II; [14]). Secondary outcomes were global severity index (GSI) on the revised symptom checklist-90 (SCL-90-R; [15]), and inventory of interpersonal problems-64 (IIP-64; [16]). Patients completed outcome measures at assessment, start of treatment, at termination, and at 1-year follow-up (see online suppl. materials at https://doi.org/10.1159/000533661 for description of outcome measures).

Statistical Procedures

Effect sizes (Cohen’s d) of symptom change during WL, treatment, and from assessment to end of treatment were calculated. We also calculated how much of the total change was achieved during WL and treatment, respectively. Minimal clinically important difference was assessed by calculating the percentage reduction of BDI-II score using 32% reduction as a cut-off to denote clinically meaningful improvement [17].

As repeated measurements on the BDI-II were nested within patients, we assessed symptom development using multilevel modeling [18]. We started with a baseline model and added one parameter at a time testing improvement in model fit for each step [19]. Model fit was estimated and compared using the −2log likelihood test [20]. First, we fitted an intercept-only model to serve as a benchmark. Then we added a random effect for the intercept and random and fixed slopes. The models were fitted with separate tests for homoscedastic and heteroscedastic variance. The slopes were estimated with linear, curvilinear, and piecewise development over time. The best model fit was achieved with a piecewise timeline model with three timelines separating the symptom slopes during waiting list, treatment, and follow-up, using fixed and random effects of intercept and time. The model was tested using unstructured covariance structure. For estimation of regression coefficients and variance components, full maximum likelihood was used [21]. After arriving at the best fitting model for time, we added use of ADM (coded 0 for nonmedicated patients and 1 for ADM users) as a predictor. Also, we added two-way interactions between ADM use and symptom change for waitlist, treatment, and follow-up. Time was coded as weeks. The first timeline was number of weeks on waiting list. The second timeline was time in active treatment (12 weeks), and the third timeline was time in follow-up (52 weeks). Time on waiting list was coded negatively, with the intercept centered at start of treatment. Supplementary analyses were conducted similarly with GSI and IIP-64.

Figure 1 shows the trajectories of symptoms for ADM and nonmedicated patients during the three time periods. During WL there was a significant and medium-sized (d = 0.468) symptom reduction for the entire sample, as well as for nonmedicated patients (d = 0.651), and ADM users (d = 0.390). WL symptom reduction constituted 35.57% of the total BDI-II change (37.89% for nonmedicated patients, 29.05% for ADM users). None of the groups achieved clinically meaningful improvement during WL. Clinically meaningful change was however achieved for all groups during treatment (see online suppl. Table S1 for demographic and clinical characteristics, S2 for BDI-II scores, and S3 for changes in BDI-II during waiting list, treatment, and the total period. Also, see online suppl. Fig. S1 for an illustration of variability in slopes).

Fig. 1.

Mean predicted BDI-II values for nonmedicated patients and ADM users at assessment, start of treatment, end of treatment, and 1-year follow-up.

Fig. 1.

Mean predicted BDI-II values for nonmedicated patients and ADM users at assessment, start of treatment, end of treatment, and 1-year follow-up.

Close modal

An independent sample t test showed there was no significant difference (t (224) = −0.997, p = 0.320) between ADM users (M = 30.08, SD = 10.273) and nonmedicated patients (M = 28.83, SD = 8.587) at assessment. Table 1 presents the results for the multilevel model analyses. There was an overall significant reduction of 0.20 BDI points per week during WL (y^ = −0.195, p = <0.001). ADM users had less symptom reduction during WL (y^ = 0.099, p = 0.039) compared to nonmedicated patients, and also had significantly higher BDI-scores (2.8) at the end of WL (note that the intercept was centered at start of treatment; y^ = 2.784, p = 0.014). There was no difference in slopes during treatment (y^ = 0.046, p = 0.693) and follow-up (y^ = −0.038, p = 0.186), indicating that the higher levels of symptoms for ADM users at the start of treatment were maintained, although both groups achieved significant effect of treatment (y^ = −0.859, p = <0.001). As for BDI-II, secondary analyses showed there was a reduction of global symptoms on GSI during WL, but there was no difference between nonmedicated patients and ADM users (see online suppl. Table S4). There was no significant reduction in IIP-64 during WL (see online suppl. Table S5).

Table 1.

Estimated values of BDI-II during waitlist, treatment, and follow-up, with ADM-group as predictor, and with interactions between ADM-group and timelines

Fixed effectsEstSEdfCIp value
Intercept 24.327 0.758 300.999 (22.835, 25.819) <0.001 
W.L. −0.195 0.035 186.028 (−0.264, −0.126) <0.001 
Treatment −0.859 0.080 279.574 (−1.015, −0.702) <0.001 
Follow-up 0.013 0.020 224.582 (−0.026, 0.052) 0.512 
ADM 2.784 1.126 301.437 (0.567, 5.000) 0.014 
W.L. * ADM 0.099 0.047 123.640 (0.005, 0.193) 0.039 
Treatment * ADM 0.046 0.117 281.079 (−0.184, 0.277) 0.693 
Follow-up * ADM −0.038 0.029 221.758 (−0.094, 0.018) 0.186 
Fixed effectsEstSEdfCIp value
Intercept 24.327 0.758 300.999 (22.835, 25.819) <0.001 
W.L. −0.195 0.035 186.028 (−0.264, −0.126) <0.001 
Treatment −0.859 0.080 279.574 (−1.015, −0.702) <0.001 
Follow-up 0.013 0.020 224.582 (−0.026, 0.052) 0.512 
ADM 2.784 1.126 301.437 (0.567, 5.000) 0.014 
W.L. * ADM 0.099 0.047 123.640 (0.005, 0.193) 0.039 
Treatment * ADM 0.046 0.117 281.079 (−0.184, 0.277) 0.693 
Follow-up * ADM −0.038 0.029 221.758 (−0.094, 0.018) 0.186 

Dependent variable is BDI-II.

Intercept centered at start of treatment. W.L. = estimated weekly change in BDI-II during waitlist; treatment = estimated weekly change in BDI-II during treatment; follow-up = estimated weekly change in BDI-II during 1-year follow-up; ADM = estimated difference in BDI-II scores at intercept between nonmedicated patients (coded 0) and users of antidepressant medication (coded 1); W.L. * ADM = interaction between weekly change in BDI-II during waitlist and use of antidepressant medication; treatment * ADM = interaction between weekly change in BDI-II during treatment and use of antidepressant medication; follow-up * ADM = interaction between weekly change in BDI-II during follow-up and use of antidepressant medication. Bold = significant at p ≤ 0.05.

BDI-II, Beck Depression Inventory-II; Est, estimated values of BDI-II; SE, standard error; df, degrees of freedom; CI, 95% confidence interval.

This study found patients with chronic depression had a significant and medium-sized (d = 0.468) symptom reduction while on WL. This is similar to the effect size estimates from other studies (d = 0.40 [4], d = 0.47 [22]). The WL reduction accounted for about 35% of the total change in the sample but did not constitute meaningful clinical change [17].

Depressed patients often experience demoralization which can be characterized as a prolonged feeling of being unable to cope and a sense of hopelessness [23]. Allostatic overload (i.e., chronic stress associated with illness becomes intolerable), and poor quality of life associated with depression may cause a state of demoralization independent of the depression and thus contribute to a worsening of total symptom load [23, 24]. Help-seeking through the act of presenting for psychotherapy suggests that individuals believe that the forthcoming treatment might provide a course of action that will be beneficial. This may facilitate hope and positive treatment expectations which play a key role in remoralization (i.e., resolving demoralization and restoring a sense of competence, mastery and well-being) [25]. Factors such as self-management and illness behavior can interact with and modify response to a specific treatment [13, 26]. Symptom reduction during WL may indicate that treatment expectations also interact with such factors modifying symptoms in the absence of treatment. This is in line with theory and research indicating treatment expectation is a primary mechanism for symptom improvement independent of the mechanisms of action in a treatment [27‒30].

Our results showed there was no difference in depression symptoms between groups at assessment. The subsequent rate of symptom improvement from assessment to the start of treatment (i.e., during WL) was significantly smaller for ADM users compared to nonmedicated patients, and at the beginning of treatment (i.e., after WL) ADM users had significantly higher levels of symptoms. This could indicate that ADM attenuates treatment expectations or increases demoralization during WL. Any type of treatment, particularly after long-term use, may increase the risk of additional problems [13]. For instance, long-term ADM-use increases the risk of coronary heart disease and cardiovascular disease [31]. Also, ADM-use may increase risk of various negative effects and reactions, which may increase chronicity and vulnerability to depressive disorders [32]. Thus, ADM users on WL may experience iatrogenic comorbidity from ADM treatment [13] that results in less symptom change compared to nonmedicated patients.

Supplementary analyses showed there was similar reduction of global symptoms as for depressive symptoms during WL, but there was no difference between nonmedicated patients and ADM users. This could indicate treatment expectation affects symptoms broadly through interaction with factors that are not disorder-specific such as self-management and illness behavior [13, 30, 33], whereas possible iatrogenic comorbidity of ADM could be specific for depression. Also, no change in IIP-64 supports findings indicating symptoms improve more quickly and strongly than interpersonal problems [34, 35]. Thus, treatment expectations during WL may help mobilize coping resources which affects symptoms related to subjective well-being, whereas additional therapeutic assistance may be required to improve interpersonal functioning [36].

A limitation to our study is that we could not randomize clients to either WL or to “no treatment” to study whether the actual waiting makes a difference compared to not receiving any treatment. Thus, confounding factors such as regression to the mean or selection bias with regard to ADM use (e.g., personality characteristics or comorbidities) could pose a threat to internal validity [37]. However, tests of group differences revealed that there were no significant differences at assessment between nonmedicated patients and ADM users on clinical and demographic variables such as presence of comorbid diagnoses, interpersonal problems, number of treatment attempts, and duration of symptoms (see online suppl. Table S1). Although lack of randomization limits causal conclusions, this provides some assurance that the difference in slopes from assessment to start of treatment could be attributed to medication use. Also, the RCT design is based on the acute disease model and evaluates therapeutic effects in untreated patients who have a recent acute onset of their disorders, whereas real-life patients often have undergone previous treatments which may have modified the course and responsiveness to subsequent treatment [38]. Thus, the generalizability of RCTs to real-world patient populations can also be problematic [39]. Another limitation of our study is that the rather strict exclusion criteria and the fact that patients were waiting for an in-patient treatment might limit the representativeness and thus the generalizability of findings. Despite the limitations, our findings corroborate previous findings that patients on WL experience a decrease in symptoms, quite likely due to positive treatment expectations.

The study was conducted ethically in accordance with the World Medical Association Declaration of Helsinki. All participants gave their written informed consent to participate in the study. The study protocol was reviewed and approved by the Regional Committees for Medical and Health Research Ethics (REC) in South East Norway (REC South East), approval numbers 2014/2355 and 2016/2003.

All authors have completed ICMJE form for disclosure of potential conflict of interest and have no conflict of interest to declare.

The study was funded by the University of Oslo and Modum Bad. The funder had no role in study design, data collection, analyses, interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.

Andreas Høstmælingen was involved in designing the study and collecting data, responsible for planning and performing the analyses, interpreting the results, and writing the report. Helene Amundsen Nissen-Lie and Bruce Wampold were involved in initiating, managing, and designing the study, involved in planning, and performing the analyses, interpreting the results, and writing the report. Pål G. Ulvenes was responsible for designing, initiating and managing the study, collecting the data, involved in planning and performing the analyses, interpreting the results, and writing the report. All authors approved the final version of the manuscript.

The data that support the findings of this study are not publicly available due to privacy or ethical restrictions. Pending approval from the treatment facility that all data are made anonymous and in compliance with GDPR and other local regulations, the data may be made available on request from the corresponding author. Further inquiries can be directed to the corresponding author.

1.
Furukawa
TA
,
Noma
H
,
Caldwell
DM
,
Honyashiki
M
,
Shinohara
K
,
Imai
H
.
Waiting list may be a nocebo condition in psychotherapy trials: a contribution from network meta-analysis
.
Acta Psychiatr Scand
.
2014
;
130
(
3
):
181
92
.
2.
Cuijpers
P
,
Karyotaki
E
,
Reijnders
M
,
Ebert
DD
.
Is psychotherapy effective? Pretending everything is fine will not help the field forward
.
Epidemiol Psychiatr Sci
.
2019
;
28
(
3
):
356
7
.
3.
Steinert
C
,
Stadter
K
,
Stark
R
,
Leichsenring
F
.
The effects of waiting for treatment: a meta-analysis of waitlist control groups in randomized controlled trials for social anxiety disorder
.
Clin Psychol Psychother
.
2017
;
24
(
3
):
649
60
.
4.
Minami
T
,
Wampold
BE
,
Serlin
RC
,
Kircher
JC
,
Brown
GS
.
Benchmarks for psychotherapy efficacy in adult major depression
.
J Consult Clin Psychol
.
2007 Apr
75
2
232
43
.
5.
Zhou
X
,
Hetrick
SE
,
Cuijpers
P
,
Qin
B
,
Barth
J
,
Whittington
CJ
.
Comparative efficacy and acceptability of psychotherapies for depression in children and adolescents: a systematic review and network meta-analysis
.
World Psychiatry
.
2015 Jun
14
2
207
22
.
6.
Michopoulos
I
,
Furukawa
TA
,
Noma
H
,
Kishimoto
S
,
Onishi
A
,
Ostinelli
EG
.
Different control conditions can produce different effect estimates in psychotherapy trials for depression
.
J Clin Epidemiol
.
2021
;
132
:
59
70
.
7.
Høstmælingen
A
,
Ulvenes
P
,
Nissen-Lie
HA
,
Eielsen
M
,
Wampold
BE
.
Comparing outcomes in chronic depression following inpatient psychotherapy for patients continuing versus discontinuing antidepressant medication
.
Clin Psychol Psychother
.
2021
;
28
(
5
):
1111
27
.
8.
Høstmælingen
A
,
Ulvenes
P
,
Nissen-Lie
HA
,
Eielsen
M
,
Wampold
BE
.
Trajectories of change in chronic depression: differences in self-criticism and somatic symptoms between users of antidepressants and nonmedicated patients
.
J Couns Psychol
.
2022
;
69
(
1
):
85
99
.
9.
Fava
GA
.
Can long-term treatment with antidepressant drugs worsen the course of depression
.
J Clin Psychiatry
.
2003 Feb
64
2
123
33
.
10.
Fava
GA
,
Offidani
E
.
The mechanisms of tolerance in antidepressant action
.
Prog Neuropsychopharmacol Biol Psychiatry
.
2011
;
35
(
7
):
1593
602
.
11.
Fava
GA
.
Rational use of antidepressant drugs
.
Psychother Psychosom
.
2014
;
83
(
4
):
197
204
.
12.
Maund
E
,
Dewar-Haggart
R
,
Williams
S
,
Bowers
H
,
Geraghty
AWA
,
Leydon
G
.
Barriers and facilitators to discontinuing antidepressant use: a systematic review and thematic synthesis
.
J Affect Disord
.
2019 Feb 15
245
38
62
.
13.
Fava
GA
,
Guidi
J
,
Rafanelli
C
,
Rickels
K
.
The clinical inadequacy of the placebo model and the development of an alternative conceptual framework
.
Psychother Psychosom
.
2017
;
86
(
6
):
332
40
.
14.
Beck
AT
,
Steer
RA
,
Brown
GK
Beck depression inventory-II
San Antonio (TX)
Psychological Corporation
1996
.
15.
Derogatis
LR
SCL-90-R. Administration, scoring and procedures manual
3rd ed.
Minneapolis (MN)
National Computer Systems
1994
.
16.
Horowitz
LM
,
Alden
LE
,
Wiggins
JS
,
Pincus
AL
Inventory of interpersonal problems manual
San Antonio (TX)
The Psychological Corporation
2000
.
17.
Button
KS
,
Kounali
D
,
Thomas
L
,
Wiles
NJ
,
Peters
TJ
,
Welton
NJ
.
Minimal clinically important difference on the Beck Depression Inventory-II according to the patient’s perspective
.
Psychol Med
.
2015 Nov
45
15
3269
79
.
18.
Raudenbush
SW
,
Bryk
AS
Hierarchical linear models: applications and data analysis methods
2nd ed.
Thousand Oaks (CA)
Sage Publications, Inc.
2002
.
19.
Singer
JD
,
Willet
JB
Applied longitudinal data analysis: modeling change and event occurrence
Oxford University Press
2003
.
20.
Fitzmaurice
GM
,
Laird
NM
,
Ware
JH
Applied longitudinal analysis
Hoboken
John Wiley & Sons
2004
.
21.
Hox
JJ
,
Moerbeek
M
,
van de Schoot
R
Multilevel analysis. Techniques and applications
3rd ed.
New York (NY)
Routledge
2018
.
22.
Hagen
R
,
Hjemdal
O
,
Solem
S
,
Kennair
LEO
,
Nordahl
HM
,
Fisher
P
.
Metacognitive therapy for depression in adults: a waiting list randomized controlled trial with six months follow-up
.
Front Psychol
.
2017
;
8
:
31
.
23.
Woźniewicz
A
,
Cosci
F
.
Clinical utility of demoralization: a systematic review of the literature
.
Clin Psychol Rev
.
2023
;
99
:
102227
.
24.
Fava
GA
,
Guidi
J
.
Clinical characterization of demoralization
.
Psychother Psychosom
.
2023
;
92
(
3
):
139
47
.
25.
Frank
JD
,
Frank
JB
Persuasion and healing: a comparative study of psychotherapy
3rd ed.
Baltimore
Johns Hopkins University Press
1991
.
26.
Benedetti
F
Placebo effects: understanding the other side of medical care
3rd ed.
Oxford
Oxford University Press
2021
.
27.
Vissers
W
,
Keijsers
GPJ
,
Kampman
M
,
Hendriks
G-J
,
Rijnders
P
,
Hutschemaekers
GJM
.
Symptom reduction without remoralization: a randomized, waiting-list controlled study aimed at separating two beneficial psychotherapy outcome effects
.
J Clin Psychol
.
2017
;
73
(
7
):
785
96
.
28.
Chavarria
V
,
Vian
J
,
Pereira
C
,
Data-Franco
J
,
Fernandes
BS
,
Berk
M
.
The placebo and nocebo phenomena: their clinical management and impact on treatment outcomes
.
Clin Ther
.
2017
;
39
(
3
):
477
86
.
29.
Constantino
MJ
,
Vîslă
A
,
Coyne
AE
,
Boswell
JF
.
A meta-analysis of the association between patients’ early treatment outcome expectation and their posttreatment outcomes
.
Psychotherapy
.
2018
;
55
(
4
):
473
85
.
30.
Wampold
BE
,
Flückiger
C
.
The alliance in mental health care: conceptualization, evidence and clinical applications
.
World Psychiatry
.
2023
;
22
(
1
):
25
41
.
31.
Bansal
N
,
Hudda
M
,
Payne
RA
,
Smith
DJ
,
Kessler
D
,
Wiles
N
.
Antidepressant use and risk of adverse outcomes: population-based cohort study
.
BJPsych Open
.
2022
;
8
(
5
):
e164
.
32.
Fava
GA
.
May antidepressant drugs worsen the conditions they are supposed to treat? The clinical foundations of the oppositional model of tolerance
.
Ther Adv Psychopharmacol
.
2020
;
10
:
2045125320970325
.
33.
Wampold
BE
,
Imel
ZE
The great psychotherapy debate. The evidence for what makes psychotherapy work
2nd ed.
New York
Routledge
2015
.
34.
Liebherz
S
,
Rabung
S
.
Do patients’ symptoms and interpersonal problems improve in psychotherapeutic hospital treatment in Germany?: a systematic review and meta-analysis
.
PLoS One
.
2014
;
9
(
8
):
e105329
.
35.
Lopes
RT
,
Gonçalves
MM
,
Fassnacht
D
,
Machado
PPP
,
Sousa
I
.
Time to improve and recover from depressive symptoms and interpersonal problems in a clinical trial
.
Clin Psychol Psychother
.
2015
;
22
(
2
):
97
105
.
36.
Howard
KI
,
Lueger
RJ
,
Maling
MS
,
Martinovich
Z
.
A phase model of psychotherapy outcome: causal mediation of change
.
J Consult Clin Psychol
.
1993
;
61
(
4
):
678
85
.
37.
Shadish
WR
,
Cook
TD
,
Campbell
DT
Experimental and quasi-experimental designs for generalized causal inference
Boston, New York
Houghton Mifflin Company
2002
.
38.
Guidi
J
,
Brakemeier
EL
,
Bockting
CLH
,
Cosci
F
,
Cuijpers
P
,
Jarrett
RB
.
Methodological recommendations for trials of psychological interventions
.
Psychother Psychosom
.
2018
;
87
(
5
):
276
84
.
39.
Rawlins
M
.
De Testimonio: on the evidence for decisions about the use of therapeutic interventions
.
Clin Med
.
2008
;
8
(
6
):
579
88
.