- Email: [email protected]
The clinical utility of different quantitative methods for measuring treatment resistance in major depression
Journal of Affective Disorders 150 (2013) 231–236
Contents lists available at ScienceDirect
Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad
Research report
The clinical utility of different quantitative methods for measuring treatment resistance in major depression Hiral Hazari a, David Christmas b,n, Keith Matthews c a
Northamptonshire Healthcare NHS Foundation Trust, Northampton, United Kingdom Advanced Interventions Service, Ninewells Hospital and Medical School, Dundee DD1 9SY, United Kingdom c Division of Neuroscience, Medical Research Institute, University of Dundee, Ninewells Hospital and Medical School, Dundee DD1 9SY, United Kingdom b
art ic l e i nf o
a b s t r a c t
Article history: Received 28 August 2012 Received in revised form 26 March 2013 Accepted 26 March 2013 Available online 11 May 2013
Background: Despite the acknowledged healthcare and economic burdens of chronic major depression, there is no agreed method to rate the degree to which patients are conceptualised as being refractory to treatment. There are a variety of tools which can be used to describe treatment resistance but their utility in clinical practice is uncertain. Methods: We used a range of contemporary tools to rate the treatment histories of patients in a variety of care settings which included: primary care; affective disorders specialist clinics; patients receiving ECT; referrals to a tertiary affective disorders service; and patients undergoing neurosurgical treatment (vagus nerve stimulation or anterior cingulotomy) for chronic, refractory major depression. Results: All tools demonstrated statistically significant differences in scores between care settings, as well as between tiers of service, although differences between some groups were small and confidence intervals were wide. The Massachusetts General Hospital staging method appeared to perform as well as more complex scoring methods and represents a reasonable compromise between time to complete and its ability to inform management decisions. Limitations: Numbers in some groups were low, but are likely to be representative. The ability of such tools to predict outcome was not examined and the proposed cut-offs require validation. Conclusions: Currently available staging methods appear to have the ability to differentiate between clinically-relevant sub-groups of patients with major depression. Further development of such tools is warranted due to their ability to not only describe characteristics of patients in different care settings, but also meet the need to have meaningful cut-offs which might guide referral to specialist treatment. & 2013 Elsevier B.V. All rights reserved.
Keywords: Treatment-refractory depression Antidepressant treatment history form Antidepressant treatment Treatment resistance
1. Introduction 1.1. Background A significant proportion of depressed patients will not respond to initial treatments (Rush et al., 2006) and the course of illness in those deemed “treatment resistant” (minimally defined as a failure to respond to at least one treatment trial) follows a particularly malignant clinical course (Fekadu et al., 2009b). Approximately 20% of patients will experience chronic depression, with symptoms lasting longer than two years (Judd et al., 1998; Kennedy et al., 2003). Importantly, there remains insufficient evidence to guide treatment choices after the first few failed antidepressant trials (Stimpson et al., 2002; Barrett et al., 2005) and clinicians struggle to predict outcome at the outset of treatment.
One of the difficulties in identifying, describing, and studying this important patient group is the lack of consensus in the definition of treatment-refractory depression (TRD). This creates a potential obstacle when identifying those who may require more intensive therapy, or who might benefit from more specialist services. Despite the challenges of chronic and treatmentrefractory MDD, there has been a persistent lack of an operational definition of ‘treatment resistance’. Similarly, it is only recently that there has been an attempt to develop standardised methods of assessing the adequacy of previous antidepressant treatments. With the increasing number of antidepressant drugs, and the complexity of combination and augmentation strategies, there is a need to adopt a reliable index of treatment adequacy which can help guide decisions about future management.
1.2. Current staging methods n
Correspondence to: Area 7, Level 6, South Block, Ninewells Hospital and Medical School, Dundee, DD1 9SY. Tel.: +44 1382 496233; fax: +44 1382 633865. E-mail address: [email protected] (D. Christmas). 0165-0327/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jad.2013.03.030
Thase and Rush (1997) initially proposed a five-stage model, based on an implied hierarchy of antidepressant classes. However,
232
H. Hazari et al. / Journal of Affective Disorders 150 (2013) 231–236
this model has several limitations (Fava, 2003): (1) A lack of a robust evidence base to support the hierarchical approach to rating (Thase et al., 2002); (2) insufficient guidance on adequate doses or durations of medication trials; (3) common approaches such as augmentation and combination are not captured; and (4) a categorical rating of treatment resistance is too imprecise for research contexts. Fava (2003) proposed the Massachusetts General Hospital staging method (MGH-S) which takes into account the number and diversity of modern antidepressant drug treatment trials and produces a continuous variable that may have greater utility in research settings. Importantly, it can score unlimited numbers of drug trials and augmentation strategies. However, it may not be flexible enough to reflect subtle variations in clinical practice. For example, ‘optimisation’ of antidepressant dose attracts the same score irrespective of the dose increase, and the MGH-S does not specify what an ‘adequate’ dose should be. Many of these limitations have been partially addressed using research-based methods of recording treatment histories in more systematic ways. One such method is the antidepressant treatment history form (ATHF) (Oquendo et al., 2002; Sackeim, 2001), which recognises the importance of dose, duration, and clinical outcome (‘response’ is regarded as an improvement of ≥50%). Each trial is assigned a score: the antidepressant resistance rating (ARR), with a score ≥3 representing the threshold for an adequate trial. A ‘confidence rating’ can also be recorded for each trial (ranging from 1–5 with 5 representing high confidence) reflecting the quality and reliability of the information used to rate each trial. The ATHF has increasingly been used in modern clinical trials to standardise criteria for inclusion and improve the reliability of medication reviews. More recent updates of the ATHF also provide ratings for psychological treatments (such as cognitive behavioural therapy and interpersonal therapy) and more experimental interventions such as vagus nerve stimulation (VNS) and transcranial magnetic stimulation (TMS). The most recently-developed tool is the Maudsley staging method (MSM) for treatment-refractory depression (Fekadu et al., 2009a). This tool places a greater emphasis on illness characteristics when rating treatment resistance, and incorporates severity and duration of illness alongside previous treatment. Preliminary studies have suggested it may have some predictive validity (Fekadu et al., 2009c), but it remains unclear if the tool is simply rating characteristics of chronic illness (which is an indicator of unfavourable prognosis) rather than predicting it. There have been a small number of reviews of staging methods and further empirical support is needed (Berlim and Turecki, 2007; Ruhé et al., 2012). Unfortunately, very few direct comparisons of these methods have been conducted. Advantages and disadvantages of different approaches are summarised in Table 1. As well as building a reliable foundation upon which to draw conclusions about previous antidepressant treatment histories, it
would be advantageous if such ratings could assist clinical recommendations and help to identify accurately those individuals who may benefit from referral to specialist affective disorder services, for example. Finally, such approaches may help to differentiate patients with ‘pseudoresistance’—those who have been undertreated over a prolonged period of time.
2. Aims and objectives The primary objective of this study was to test the face validity of three different methods (of varying complexity) for assessing treatment resistance in different populations with major depression. Second, the study intended to assess the clinical utility of these methods to discriminate between different tiers of treatment. Finally, the study aimed to determine if more complex ways of scoring treatment resistance conferred any advantages over shorter, simpler methods from a research perspective.
3. Study method 3.1. Subjects Data on previous antidepressant treatments were obtained from the following patient groups, listed in order of presumed increasing treatment resistance. As this was a preliminary study, sample sizes were pragmatic and were limited by time constraints. All data were pseudo-anonymised. Case selection was based on the random sampling procedure in the statistical package. (1) Primary care (N¼14): We approached a local academic research group who held a dataset containing information on primary care patients with diagnoses of major depression, including demographic details along with a record of current and previous antidepressant treatment history (N¼95). Cases were randomly selected. (2) Dundee affective disorders clinic (ADC) (N ¼11): This clinic ran between 1998 and 2003, providing protocolised drug treatment for patients with a primary diagnosis of major depression patients referred from both primary and secondary care. Cases were randomly-selected from the total dataset (N¼920) and treatment histories were reviewed. (3) Patients undergoing ECT in secondary care (N¼10): A clinical database containing systematically-collected information on patients receiving ECT at a local psychiatric hospital was used. Cases were randomly-selected from the dataset (N¼30) and treatment histories were completed from the case notes.
Table 1 A comparison of different staging methods of treatment-resistance. Time required for completion
TR-S Short
MGH-S Medium
MSM Medium
ATHF Long
Detail needed on previous trials
Minimal—simple exposure to a trial of an ‘adequate’ dose and duration
Moderate—requires details on optimisation, combination, and augmentation
Moderate—characteristics of illness and course are needed to score
Scoring
Ordinal: 1–5
Scale (no upper limit)
Scale (maximum score of 17)
Presumed/ intended use
Brief categorical indicator of staging
More detailed assessment of previous treatment exposure
Clinical practice/specialist affective disorders services.
High—each trial needs detail of dose, duration, and reliability of information. Majority of drugs are potentially recordable Scale—each trial is scored, and total number of trials can be summed. No upper limit Research settings; used to determine adequacy of each treatment trial
TR-S: Thase and Rush staging; MGH-S: Massachusetts General Hospital staging method; MSM: Maudsley staging method; ATHF: Antidepressant treatment history form.
H. Hazari et al. / Journal of Affective Disorders 150 (2013) 231–236
(4) Tertiary care referrals (N¼34): Consecutive referrals (between October 2006 and November 2010) to the Advanced Interventions Service in Dundee for advice on further pharmacological and psychological treatment options for treatment-resistant/chronic depression had data on antidepressant treatment histories recorded. Those that did not proceed to neurosurgical treatment, and where the diagnosis was major depression were included. (5) Vagus nerve stimulation (VNS) (N ¼16): Consecutive individuals treated with vagus nerve stimulation for chronic, refractory major depression in Dundee between 2002 and 2011 were included from a pre-existing dataset. (6) Neurosurgical treatment (anterior capsulotomy/anterior cingulotomy) (N¼17): Consecutive individuals who were treated by an ablative neurosurgical procedure for chronic, refractory major depression in Dundee since 1992 were included from a pre-existing dataset. Exclusion criteria applied prior to random sampling (where applicable) were: (a) a primary diagnosis other than major depression; and (b) antidepressant prescriptions for nondepressive illness. All patients had received a diagnosis of depressive episode by a psychiatrist using ICD-10 criteria (World Health Organisation, 1992) except the primary care group, where the diagnosis had been made by general practitioners based on clinical interview. Retrospective, operationalised ICD-10 and DSM-IV diagnoses were also confirmed using OPCRIT (McGuffin et al., 1991; Williams et al., 1996).
233
gives greater weight to treatment trials with better supporting evidence. c) The treatment precision index (TPI) was calculated. This is the ratio of the sum of ATHF scores of adequate trials to the sum of ATHF scores of all trials. The ratio represents the proportion of the total score that is made up of adequate trials. A higher number indicates a higher proportion of adequate trials. The maximum score is 1 (all trials are rated adequate). It is a global indicator of treatment robustness. d) The DATA score was given by the following formula: (MGH staging score+(Thase and Rush stage 3)+composite score for all treatments) TPI
3.3. Statistical analysis Statistical analysis was performed using SPSS Version 16.0 for Windows (SPSS, 2007). Categorical data were compared using the Chi-squared test. For normally distributed data, comparison between two groups was performed with independent samples t-tests. Where there were three or more categories, one-way ANOVA was used. To examine for differences between two variables (or categories) that were not normally distributed, the Wilcoxon/Mann–Whitney U test was used. For three or more categories, the Kruskal–Wallis test was used. 3.4. Ethical considerations Caldicott Guardian approval was obtained to authorise access and to hold patient data. All data were pseudo-anonymised, and permission to access the primary care database was obtained from the general practitioner research team.
3.2. Assessments The required information for scoring on each scale was extracted from clinical case notes in all patient groups using a standard proforma. Treatment histories included details of all psychotropic medications (doses and durations) during the current major depressive episode, along with courses of ECT (where relevant), and confidence ratings for the ATHF. In addition to scores on the TR-S, MGH-S, and ATHF, an additional score was calculated. The ‘Dundee antidepressant treatment adequacy’ (DATA) score was developed as a more complex metric than other staging methods and it attempts to compensate for some of the limitations of other methods. The aim was to retain a continuous variable which takes into account the level of treatment resistance along with the adequacy of past treatments. It was calculated using the following formula: a) First, Thase and Rush stage (an integer between 0 and 5) was multiplied by 3 to make it the same order of magnitude as the MGH staging method score (typically a decimal between 0 and 15 for most populations, but it can be higher). b) Composite scores (the product of the ARR from the ATHF and the confidence rating) were then calculated for each trial. This
4. Results A comparison between the groups and methods is shown below in Table 2. This reveals low levels of pharmacological treatment of depression in primary care with the mean number of adequate antidepressant trials according to the ATHF being less than one. This may, of course, reflect poor recording of antidepressant trials, inadequate treatment, or both. The identical scores between the TR-S and ATHF in the primary care sample reflect the categorical nature of the TR-S. At low levels of treatment resistance, there will be a high correlation between each adequate trial on the ATHF and the corresponding TR-S stage. One-way ANOVA was used to assess the difference between mean staging method scores across the six patient groups. This revealed highly significant differences between groups for the TR-S (F(5,95) ¼31.475), ATHF number of adequate treatments (F(5,95 ¼ 32.660), MGH-S (F(5,95) ¼39.561), and the DATA Score (F (5,95) ¼40.229). All results were statistically significant at the Po 0.001 level.
Table 2 Mean staging method scores (95% confidence interval). Primary care (N ¼ 14) TR-S 0.43 MGH-S 0.54 No. of ‘adequate’ treatments (ATHF) 0.43 DATA 3.69
ADC (N ¼ 10)
(0.06–0.80) 1.80 (0.02–1.05) 5.10 (0.06–0.80) 3.20 (0.53–6.86) 22.73
ECT (N ¼ 10)
(0.99–2.61) 2.00 (2.46–7.74) 8.70 (1.55–4.85) 5.60 (13.41–32.05) 35.22
Tertiary care (N ¼ 34) VNS (N ¼ 16)
(1.11–2.89) 3.21 (2.76–3.65) (6.49–10.91) 10.81 (9.36–12.26) (3.60–7.60) 8.41 (7.16–9.66) (24.70–45.73) 49.08 (42.45–55.70)
3.62 13.25 9.19 55.71
Neurosurgery (N ¼ 17)
(2.93–4.32) 4.41 (3.93–4.89) (11.02–15.48) 15.56 (14.27–16.84) (7.14–11.24) 12.06 (10.77–13.35) (46.41–65.02) 71.30 (64.87–77.74)
TR-S: Thase and Rush staging; MGH-S: Massachusetts General Hospital staging method; MSM: Maudsley staging method; ATHF: Antidepressant treatment history form; DATA: Dundee antidepressant treatment adequacy score.
234
H. Hazari et al. / Journal of Affective Disorders 150 (2013) 231–236
Table 3 Mean staging method scores between tiers (95% confidence interval).
TR-S MGH-S No. of ‘adequate’ treatments (ATHF) DATA
Primary care (N ¼ 14)
Secondary care (N ¼20)
Tertiary care (N ¼ 67)
0.43 0.54 0.43 3.69
1.90 6.90 4.40 28.97
3.61 12.60 9.52 56.30
(0.06–0.80) (0.02–1.05) (0.06–0.80) (0.53–6.86)
(1.35–2.45) (5.13–8.67) (3.10–5.70) (21.97–35.98)
(3.30–3.93) (11.56–13.64) (8.62–10.42) (51.59–61.01)
TR-S: Thase and Rush staging; MGH-S: Massachusetts General Hospital staging method; MSM: Maudsley staging method; ATHF: Antidepressant treatment history form; DATA: Dundee antidepressant treatment adequacy score.
Table 4 Putative cut-offs for secondary and tertiary care based on MGH-S and ATHF scores.
MGH-S ATHF no. of adequate treatments
Secondary care
Tertiary care
4.00 2.25
9.00 7.00
There is, however, a considerable degree of overlap between the confidence intervals of each of these groups, which is to be expected given the heterogeneity of such groups. This overlap may limit the ability of each staging method to discriminate between levels which are often, to some extent, artificial. Being able to distinguish between primary and secondary care, and between secondary and tertiary care may be more useful in everyday clinical practice. Therefore, these categories were used to compare patient groups. Secondary care patients were constituted from ADC patients and ECT clinic patients (N ¼20), and tertiary care patients were made up of tertiary care referrals, VNS patients, and neurosurgery patients (N ¼ 67). Results are shown below in Table 3. Comparisons on the MGH-S were made between each tier were using t-tests. The difference between each tier (primary/secondary and secondary/tertiary) was statistically significant (Po 0.001) for each measure used Table 4. MGH-S scores for these groups are shown below in Fig. 1.
5. Discussion 5.1. Conclusions Whilst it is challenging to numerically quantify the degree of antidepressant treatment resistance, these data suggest that currently available staging methods may have some capacity to differentiate between clinically-relevant patient groups. By ‘benchmarking’ the levels of treatment received in specific groups, it may be helpful to identify those patients who would benefit from either specific interventions (e.g. ECT) or from specialist clinical review. In Scotland, for example, the standards for integrated care pathways (ICPs) require that specialist assessment should be available for those with treatment-refractory depression (NHS QIS, 2007). Being able to identify those patients who would benefit most would assist services in identifying patients within different parts of the existing care pathway. In addition, such a rating might also highlight those individuals whose previous treatments would prompt the treating clinician to review management in more detail. Whilst the MGH-S and DATA scores would appear to have an equally satisfactory ability to differentiate between different treatment groups (and would appear more sensitive than the TR-S), the relative ease of completion for the MGH-S method would make this a preferable choice for routine clinical practice. Indeed, the advantages of more complex methods of assessing treatment
resistance are less clear when the additional efforts required to calculate them are considered. Whilst ATHF scores (and the derived DATA scores) may have a role in research settings and for defining inclusion criteria for clinical trials, the MGH-S would appear to possess the ability to differentiate between different levels of resistance, as well as providing a continuous variable that is responsive to further treatment. Indeed, we have yet to identify ceiling effects using the MGH-S or DATA scores. The TR-S, when being used to rate low numbers of treatment trials, may be relatively insensitive to potentially therapeutic optimisations of treatment due to its categorical nature. The MSM was not included in this study but further assessment of this tool is planned. We have attempted to provide putative cut-off points at which it would be reasonable to conclude that the patient’s level of resistance is characteristic of the next tier. For each scale, the 25th percentile was considered appropriate since 75% of patients in that tier would be above this score. Such ‘cut-offs’ are shown in Table 4. These figures would suggest that referral to secondary care is occurring after two failed antidepressant trials – this is consistent with the guidelines of NICE (2009) – and referral to tertiary care after seven failed trials (including augmentation strategies). In addition, based on this (small) group of patients, ECT is being delivered to patients who have had four failed antidepressant trials and who have an MGH-S score of 6.0. It is notable that levels of treatment resistance were comparatively low in the primary care sample. However, our findings are consistent with other ratings of treatment in primary care patients with depression. For example, Smith et al. (2011) reported median MGH-S scores of 1.5, reflecting the equivalent of one fullyoptimised antidepressant trial in a population where chronic depression was present in 56%. Even in a US academic psychiatry service, Petersen et al. (2005) found that scores on the MGH-S and TR-S scales were relatively low (the mean 7SD of the entire sample on the MGH-S was 1.6 71.2). 5.2. Limitations This study is the first to compare different methods for describing/staging treatment-refractory depression across such a wide range of illness histories. Treatment histories were based on case note reviews in all cases (thereby avoiding problems due to recall bias) so the likelihood that previous treatments were missed is reduced. Further, the sample is likely to be representative of a real-world clinical population as they were randomly selected from target populations relevant to clinical practice. However, there are weaknesses associated with a study such as this. The numbers are small, and these findings should be considered preliminary. Although attempts were made to ensure that previous treatment trials were not overlooked, the possibility that low levels of treatment in the primary care group reflected insufficient documentation cannot be completely excluded. Further, without being able to corroborate every treatment trial with the patient, it is not possible to always conclude that medication was taken exactly as documented. Clinical correspondence for all groups may, in some cases, overestimate the
H. Hazari et al. / Journal of Affective Disorders 150 (2013) 231–236
235
Fig. 1. Box plot displaying MGH-S Scores for three tiers of service.
robustness of a medication trial. In most cases, assessment of treatment adequacy requires a systematic process of case note review. Assessment of adequacy of previous psychological therapies is an essential process but is complex and would benefit from reliable tools. We agree with Ruhé et al. (2012) that detailed demographics and illness characteristics are preferable in such studies. However, it was not possible to obtain these for all groups in this report; although they are available for the patients in tertiary care and above. We hope to report on these in due course. Using such rating scores to rigidly define the boundaries of clinical services is likely to be premature. It can be seen in Fig. 1 that some patients in the secondary care group had resistance levels comparable with those patients being seen in a tertiary care service. This is likely to reflect relatively resistant patients remaining in secondary care and some patients with lesser degrees of resistance being referred to tertiary care. Social class effects influence waiting times for some surgical interventions (Pell et al., 2000) and may influence referral rates in psychiatry. In the real world, it might be difficult to eliminate crossover; particularly when illness characteristics (e.g. symptom severity, levels of distress, and risk factors such as suicidality) and patient factors (e.g. motivation) will affect referral rates. It is anticipated that larger sample sizes are needed to confirm that these findings are robust across all treatment groups, and further work is required to determine if these ratings can predict outcome. Of course, those with more refractory depression are, by definition, less likely to respond to treatment. A less favourable illness outcome is associated with failure to respond to two or
more drugs and predicting outcome based on indices that already reflect a poor-prognosis group will be challenging. However, it should be possible to determine if response to ECT, for example, varies according to the level of treatment resistance prior to treatment. Currently, research on this prediction is equivocal (Prudic et al., 1990; Kindler et al., 1991). Such a finding would have implications for treatment decisions. Finally, it would be interesting to determine if those patients above a certain threshold who are then referred to more specialist services do better than those who are not referred. Consequently, we would encourage specialist affective disorders services to systematise data collection to aid comparisons of patients and outcomes between services. Our results provide putative support for the use of the MGH-S as a useful research tool.
Role of funding source None.
Conflict of interest HH has received travel and conference attendance from Janssen-Cilag and has had sponsorship for training courses from Wyeth. DC provides clinical management for the Dundee Advanced Interventions Service. He has received consultancy fees from Servier Laboratories and honoraria for lectures from Wyeth and Lilly. He has received travel and conference attendance from Medtronic and Cyberonics Inc. KM has Chaired advisory boards for studies of Deep Brain Stimulation for Obsessive–Compulsive Disorder sponsored by Medtronic. He has received educational grants from Cyberonics Inc & Schering Plough, and he has received research funding from St Jude Medical for a multi-centre clinical trial of deep brain stimulation for depression. He has received travel and accommodation support to attend meetings from Medtronic and St Jude Medical.
236
H. Hazari et al. / Journal of Affective Disorders 150 (2013) 231–236
Acknowledgements We are grateful to primary and secondary care colleagues who allowed us access to clinical records in order to complete the study.
References Barrett, B., Byford, S., Knapp, M., 2005. Evidence of cost-effective treatments for depression: a systematic review. Journal of Affective Disorders 84, 1–13. Berlim, M.T., Turecki, G., 2007. Definition, assessment, and staging of treatmentresistant refractory major depression: a review of current concepts and methods. Canadian Journal of Psychiatry 52, 46–54. Fava, M., 2003. Diagnosis and definition of treatment-resistant depression. Biological Psychiatry 53, 649–659. Fekadu, A., Wooderson, S., Donaldson, C., Markopoulou, K., Masterson, B., Poon, L., Cleare, A.J., 2009a. A multidimensional tool to quantify treatment resistance in depression: the Maudsley staging method. Journal of Clinical Psychiatry 70, 177–184. Fekadu, A., Wooderson, S.C., Markopoulo, K., Donaldson, C., Papadopoulos, A., Cleare, A.J., 2009b. What happens to patients with treatment-resistant depression? A systematic review of medium to long term outcome studies. Journal of Affective Disorders 116, 4–11. Fekadu, A., Wooderson, S.C., Markopoulou, K., Cleare, A.J., 2009c. The Maudsley staging method for treatment-resistant depression: prediction of longer-term outcome and persistence of symptoms. Journal of Clinical Psychiatry 70, 952–957. Judd, L.L., Akiskal, H.S., Maser, J.D., Zeller, P.J., Endicott, J., Coryell, W., Paulus, M.P., Kunovac, J.L., Leon, A.C., Mueller, T.I., Rice, J.A., Keller, M.B., 1998. A prospective 12-year study of subsyndromal and syndromal depressive symptoms in unipolar major depressive disorders. Archives of General Psychiatry 55, 694–700. Kennedy, N., Abbott, R., Paykel, E.S., 2003. Remission and recurrence of depression in the maintenance era: long-term outcome in a Cambridge cohort. Psychological Medicine 33, 827–838. Kindler, S., Shapira, B., Hadjez, J., Abramowitz, M., Brom, D., Lerer, B., 1991. Factors influencing response to bilateral electroconvulsive therapy in major depression. Convulsive Therapy 7, 245–254. McGuffin, P., Farmer, A., Harvey, I., 1991. A polydiagnostic application of operational criteria in studies of psychotic illness. Development and reliability of the OPCRIT system. Archives of General Psychiatry 48, 764–770. National Institute for Health and Clinical Excellence (NICE), 2009. Depression: the Treatment and Management of Depression in Adults. National Collaborating Centre for Mental Health, London. NHS QIS, 2007. Standards for Integrated Care Pathways for Mental Health. NHS Quality Improvement Scotland, Edinburgh.
Oquendo, M.A., Kamali, M., Ellis, S.P., Grunebaum, M.F., Malone, K.M., Brodsky, B.S., Sackeim, H.A., Mann, J.J., 2002. Adequacy of antidepressant treatment after discharge and the occurrence of suicidal acts in major depression: a prospective study. American Journal of Psychiatry 159, 1746–1751. Pell, J.P., Pell, A.C.H., Norrie, J., Ford, I., Cobbe, S.M., Hart, J.T., 2000. Effect of socioeconomic deprivation on waiting time for cardiac surgery: retrospective cohort study. British Medical Journal 320, 15–19. Petersen, T., Papakostas, G.I., Posternak, M.A., Kant, A., Guyker, W.M., Iosifescu, D.V., Yeung, A.S., Nierenberg, A.A., Fava, M., 2005. Empirical testing of two models for staging antidepressant treatment resistance. Journal of Clinical Psychopharmacology 25, 336–341. Prudic, J., Sackeim, H.A., Devanand, D.P., 1990. Medication resistance and clinical response to electroconvulsive therapy. Psychiatry Research 31, 287–296. Ruhé, H.G., van Rooijen, G., Spijker, J., Peeters, F.P.M.L., Schene, A.H., 2012. Staging methods for treatment resistant depression. A systematic review. Journal of Affective Disorders 137, 35–45. Rush, A.J., Trivedi, M.H., Wisniewski, S.R., Nierenberg, A.A., Stewart, J.W., Warden, D., Niederehe, G., Thase, M.E., Lavori, P.W., Lebowitz, B.D., McGrath, P.J., Rosenbaum, J. F., Sackeim, H.A., Kupfer, D.J., Luther, J., Fava, M., 2006. Acute and longer-term outcomes in depressed outpatients requiring one or several treatment steps: a STARnD report. American Journal of Psychiatry 163, 1905–1917. Sackeim, H.A., 2001. The definition and meaning of treatment-resistant depression. Journal of Clinical Psychiatry 62 (Suppl. 16), 10–17. Smith, D.J., Griffiths, E., Kelly, M., Hood, K., Craddock, N., Simpson, S.A., 2011. Unrecognised bipolar disorder in primary care patients with depression. British Journal of Psychiatry 199, 49–56. SPSS, 2007. Version 16.0.2 for Microsoft Windowss. SPSS Inc, Chicago, IL. Stimpson, N., Agrawal, N., Lewis, G., 2002. Randomised controlled trials investigating pharmacological and psychological interventions for treatment-refractory depression. Systematic review. British Journal of Psychiatry 181, 284–294. Thase, M.E., Rush, A.J., 1997. When at first you don't succeed: sequential strategies for antidepressant nonresponders. Journal of Clinical Psychiatry 58 (Suppl 13), 23–29. Thase, M.E., Rush, A.J., Howland, R.H., Kornstein, S.G., Kocsis, J.H., Gelenberg, A.J., Schatzberg, A.F., Koran, L.M., Keller, M.B., Russell, J.M., Hirschfeld, R.M., LaVange, L.M., Klein, D.N., Fawcett, J., Harrison, W., 2002. Double-blind switch study of imipramine or sertraline treatment of antidepressant-resistant chronic depression. Archives of General Psychiatry 59, 233–239. Williams, J., Farmer, A.E., Ackenheil, M., Kaufmann, C.A., McGuffin, P., 1996. A multicentre inter-rater reliability study using the OPCRIT computerized diagnostic system. Psychological Medicine 26, 775–783. World Health Organisation, 1992. International Statistical Classification of Diseases and Related Health Problems, 10th Revision. World Health Organisation, Geneva.
Contents lists available at ScienceDirect
Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad
Research report
The clinical utility of different quantitative methods for measuring treatment resistance in major depression Hiral Hazari a, David Christmas b,n, Keith Matthews c a
Northamptonshire Healthcare NHS Foundation Trust, Northampton, United Kingdom Advanced Interventions Service, Ninewells Hospital and Medical School, Dundee DD1 9SY, United Kingdom c Division of Neuroscience, Medical Research Institute, University of Dundee, Ninewells Hospital and Medical School, Dundee DD1 9SY, United Kingdom b
art ic l e i nf o
a b s t r a c t
Article history: Received 28 August 2012 Received in revised form 26 March 2013 Accepted 26 March 2013 Available online 11 May 2013
Background: Despite the acknowledged healthcare and economic burdens of chronic major depression, there is no agreed method to rate the degree to which patients are conceptualised as being refractory to treatment. There are a variety of tools which can be used to describe treatment resistance but their utility in clinical practice is uncertain. Methods: We used a range of contemporary tools to rate the treatment histories of patients in a variety of care settings which included: primary care; affective disorders specialist clinics; patients receiving ECT; referrals to a tertiary affective disorders service; and patients undergoing neurosurgical treatment (vagus nerve stimulation or anterior cingulotomy) for chronic, refractory major depression. Results: All tools demonstrated statistically significant differences in scores between care settings, as well as between tiers of service, although differences between some groups were small and confidence intervals were wide. The Massachusetts General Hospital staging method appeared to perform as well as more complex scoring methods and represents a reasonable compromise between time to complete and its ability to inform management decisions. Limitations: Numbers in some groups were low, but are likely to be representative. The ability of such tools to predict outcome was not examined and the proposed cut-offs require validation. Conclusions: Currently available staging methods appear to have the ability to differentiate between clinically-relevant sub-groups of patients with major depression. Further development of such tools is warranted due to their ability to not only describe characteristics of patients in different care settings, but also meet the need to have meaningful cut-offs which might guide referral to specialist treatment. & 2013 Elsevier B.V. All rights reserved.
Keywords: Treatment-refractory depression Antidepressant treatment history form Antidepressant treatment Treatment resistance
1. Introduction 1.1. Background A significant proportion of depressed patients will not respond to initial treatments (Rush et al., 2006) and the course of illness in those deemed “treatment resistant” (minimally defined as a failure to respond to at least one treatment trial) follows a particularly malignant clinical course (Fekadu et al., 2009b). Approximately 20% of patients will experience chronic depression, with symptoms lasting longer than two years (Judd et al., 1998; Kennedy et al., 2003). Importantly, there remains insufficient evidence to guide treatment choices after the first few failed antidepressant trials (Stimpson et al., 2002; Barrett et al., 2005) and clinicians struggle to predict outcome at the outset of treatment.
One of the difficulties in identifying, describing, and studying this important patient group is the lack of consensus in the definition of treatment-refractory depression (TRD). This creates a potential obstacle when identifying those who may require more intensive therapy, or who might benefit from more specialist services. Despite the challenges of chronic and treatmentrefractory MDD, there has been a persistent lack of an operational definition of ‘treatment resistance’. Similarly, it is only recently that there has been an attempt to develop standardised methods of assessing the adequacy of previous antidepressant treatments. With the increasing number of antidepressant drugs, and the complexity of combination and augmentation strategies, there is a need to adopt a reliable index of treatment adequacy which can help guide decisions about future management.
1.2. Current staging methods n
Correspondence to: Area 7, Level 6, South Block, Ninewells Hospital and Medical School, Dundee, DD1 9SY. Tel.: +44 1382 496233; fax: +44 1382 633865. E-mail address: [email protected] (D. Christmas). 0165-0327/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jad.2013.03.030
Thase and Rush (1997) initially proposed a five-stage model, based on an implied hierarchy of antidepressant classes. However,
232
H. Hazari et al. / Journal of Affective Disorders 150 (2013) 231–236
this model has several limitations (Fava, 2003): (1) A lack of a robust evidence base to support the hierarchical approach to rating (Thase et al., 2002); (2) insufficient guidance on adequate doses or durations of medication trials; (3) common approaches such as augmentation and combination are not captured; and (4) a categorical rating of treatment resistance is too imprecise for research contexts. Fava (2003) proposed the Massachusetts General Hospital staging method (MGH-S) which takes into account the number and diversity of modern antidepressant drug treatment trials and produces a continuous variable that may have greater utility in research settings. Importantly, it can score unlimited numbers of drug trials and augmentation strategies. However, it may not be flexible enough to reflect subtle variations in clinical practice. For example, ‘optimisation’ of antidepressant dose attracts the same score irrespective of the dose increase, and the MGH-S does not specify what an ‘adequate’ dose should be. Many of these limitations have been partially addressed using research-based methods of recording treatment histories in more systematic ways. One such method is the antidepressant treatment history form (ATHF) (Oquendo et al., 2002; Sackeim, 2001), which recognises the importance of dose, duration, and clinical outcome (‘response’ is regarded as an improvement of ≥50%). Each trial is assigned a score: the antidepressant resistance rating (ARR), with a score ≥3 representing the threshold for an adequate trial. A ‘confidence rating’ can also be recorded for each trial (ranging from 1–5 with 5 representing high confidence) reflecting the quality and reliability of the information used to rate each trial. The ATHF has increasingly been used in modern clinical trials to standardise criteria for inclusion and improve the reliability of medication reviews. More recent updates of the ATHF also provide ratings for psychological treatments (such as cognitive behavioural therapy and interpersonal therapy) and more experimental interventions such as vagus nerve stimulation (VNS) and transcranial magnetic stimulation (TMS). The most recently-developed tool is the Maudsley staging method (MSM) for treatment-refractory depression (Fekadu et al., 2009a). This tool places a greater emphasis on illness characteristics when rating treatment resistance, and incorporates severity and duration of illness alongside previous treatment. Preliminary studies have suggested it may have some predictive validity (Fekadu et al., 2009c), but it remains unclear if the tool is simply rating characteristics of chronic illness (which is an indicator of unfavourable prognosis) rather than predicting it. There have been a small number of reviews of staging methods and further empirical support is needed (Berlim and Turecki, 2007; Ruhé et al., 2012). Unfortunately, very few direct comparisons of these methods have been conducted. Advantages and disadvantages of different approaches are summarised in Table 1. As well as building a reliable foundation upon which to draw conclusions about previous antidepressant treatment histories, it
would be advantageous if such ratings could assist clinical recommendations and help to identify accurately those individuals who may benefit from referral to specialist affective disorder services, for example. Finally, such approaches may help to differentiate patients with ‘pseudoresistance’—those who have been undertreated over a prolonged period of time.
2. Aims and objectives The primary objective of this study was to test the face validity of three different methods (of varying complexity) for assessing treatment resistance in different populations with major depression. Second, the study intended to assess the clinical utility of these methods to discriminate between different tiers of treatment. Finally, the study aimed to determine if more complex ways of scoring treatment resistance conferred any advantages over shorter, simpler methods from a research perspective.
3. Study method 3.1. Subjects Data on previous antidepressant treatments were obtained from the following patient groups, listed in order of presumed increasing treatment resistance. As this was a preliminary study, sample sizes were pragmatic and were limited by time constraints. All data were pseudo-anonymised. Case selection was based on the random sampling procedure in the statistical package. (1) Primary care (N¼14): We approached a local academic research group who held a dataset containing information on primary care patients with diagnoses of major depression, including demographic details along with a record of current and previous antidepressant treatment history (N¼95). Cases were randomly selected. (2) Dundee affective disorders clinic (ADC) (N ¼11): This clinic ran between 1998 and 2003, providing protocolised drug treatment for patients with a primary diagnosis of major depression patients referred from both primary and secondary care. Cases were randomly-selected from the total dataset (N¼920) and treatment histories were reviewed. (3) Patients undergoing ECT in secondary care (N¼10): A clinical database containing systematically-collected information on patients receiving ECT at a local psychiatric hospital was used. Cases were randomly-selected from the dataset (N¼30) and treatment histories were completed from the case notes.
Table 1 A comparison of different staging methods of treatment-resistance. Time required for completion
TR-S Short
MGH-S Medium
MSM Medium
ATHF Long
Detail needed on previous trials
Minimal—simple exposure to a trial of an ‘adequate’ dose and duration
Moderate—requires details on optimisation, combination, and augmentation
Moderate—characteristics of illness and course are needed to score
Scoring
Ordinal: 1–5
Scale (no upper limit)
Scale (maximum score of 17)
Presumed/ intended use
Brief categorical indicator of staging
More detailed assessment of previous treatment exposure
Clinical practice/specialist affective disorders services.
High—each trial needs detail of dose, duration, and reliability of information. Majority of drugs are potentially recordable Scale—each trial is scored, and total number of trials can be summed. No upper limit Research settings; used to determine adequacy of each treatment trial
TR-S: Thase and Rush staging; MGH-S: Massachusetts General Hospital staging method; MSM: Maudsley staging method; ATHF: Antidepressant treatment history form.
H. Hazari et al. / Journal of Affective Disorders 150 (2013) 231–236
(4) Tertiary care referrals (N¼34): Consecutive referrals (between October 2006 and November 2010) to the Advanced Interventions Service in Dundee for advice on further pharmacological and psychological treatment options for treatment-resistant/chronic depression had data on antidepressant treatment histories recorded. Those that did not proceed to neurosurgical treatment, and where the diagnosis was major depression were included. (5) Vagus nerve stimulation (VNS) (N ¼16): Consecutive individuals treated with vagus nerve stimulation for chronic, refractory major depression in Dundee between 2002 and 2011 were included from a pre-existing dataset. (6) Neurosurgical treatment (anterior capsulotomy/anterior cingulotomy) (N¼17): Consecutive individuals who were treated by an ablative neurosurgical procedure for chronic, refractory major depression in Dundee since 1992 were included from a pre-existing dataset. Exclusion criteria applied prior to random sampling (where applicable) were: (a) a primary diagnosis other than major depression; and (b) antidepressant prescriptions for nondepressive illness. All patients had received a diagnosis of depressive episode by a psychiatrist using ICD-10 criteria (World Health Organisation, 1992) except the primary care group, where the diagnosis had been made by general practitioners based on clinical interview. Retrospective, operationalised ICD-10 and DSM-IV diagnoses were also confirmed using OPCRIT (McGuffin et al., 1991; Williams et al., 1996).
233
gives greater weight to treatment trials with better supporting evidence. c) The treatment precision index (TPI) was calculated. This is the ratio of the sum of ATHF scores of adequate trials to the sum of ATHF scores of all trials. The ratio represents the proportion of the total score that is made up of adequate trials. A higher number indicates a higher proportion of adequate trials. The maximum score is 1 (all trials are rated adequate). It is a global indicator of treatment robustness. d) The DATA score was given by the following formula: (MGH staging score+(Thase and Rush stage 3)+composite score for all treatments) TPI
3.3. Statistical analysis Statistical analysis was performed using SPSS Version 16.0 for Windows (SPSS, 2007). Categorical data were compared using the Chi-squared test. For normally distributed data, comparison between two groups was performed with independent samples t-tests. Where there were three or more categories, one-way ANOVA was used. To examine for differences between two variables (or categories) that were not normally distributed, the Wilcoxon/Mann–Whitney U test was used. For three or more categories, the Kruskal–Wallis test was used. 3.4. Ethical considerations Caldicott Guardian approval was obtained to authorise access and to hold patient data. All data were pseudo-anonymised, and permission to access the primary care database was obtained from the general practitioner research team.
3.2. Assessments The required information for scoring on each scale was extracted from clinical case notes in all patient groups using a standard proforma. Treatment histories included details of all psychotropic medications (doses and durations) during the current major depressive episode, along with courses of ECT (where relevant), and confidence ratings for the ATHF. In addition to scores on the TR-S, MGH-S, and ATHF, an additional score was calculated. The ‘Dundee antidepressant treatment adequacy’ (DATA) score was developed as a more complex metric than other staging methods and it attempts to compensate for some of the limitations of other methods. The aim was to retain a continuous variable which takes into account the level of treatment resistance along with the adequacy of past treatments. It was calculated using the following formula: a) First, Thase and Rush stage (an integer between 0 and 5) was multiplied by 3 to make it the same order of magnitude as the MGH staging method score (typically a decimal between 0 and 15 for most populations, but it can be higher). b) Composite scores (the product of the ARR from the ATHF and the confidence rating) were then calculated for each trial. This
4. Results A comparison between the groups and methods is shown below in Table 2. This reveals low levels of pharmacological treatment of depression in primary care with the mean number of adequate antidepressant trials according to the ATHF being less than one. This may, of course, reflect poor recording of antidepressant trials, inadequate treatment, or both. The identical scores between the TR-S and ATHF in the primary care sample reflect the categorical nature of the TR-S. At low levels of treatment resistance, there will be a high correlation between each adequate trial on the ATHF and the corresponding TR-S stage. One-way ANOVA was used to assess the difference between mean staging method scores across the six patient groups. This revealed highly significant differences between groups for the TR-S (F(5,95) ¼31.475), ATHF number of adequate treatments (F(5,95 ¼ 32.660), MGH-S (F(5,95) ¼39.561), and the DATA Score (F (5,95) ¼40.229). All results were statistically significant at the Po 0.001 level.
Table 2 Mean staging method scores (95% confidence interval). Primary care (N ¼ 14) TR-S 0.43 MGH-S 0.54 No. of ‘adequate’ treatments (ATHF) 0.43 DATA 3.69
ADC (N ¼ 10)
(0.06–0.80) 1.80 (0.02–1.05) 5.10 (0.06–0.80) 3.20 (0.53–6.86) 22.73
ECT (N ¼ 10)
(0.99–2.61) 2.00 (2.46–7.74) 8.70 (1.55–4.85) 5.60 (13.41–32.05) 35.22
Tertiary care (N ¼ 34) VNS (N ¼ 16)
(1.11–2.89) 3.21 (2.76–3.65) (6.49–10.91) 10.81 (9.36–12.26) (3.60–7.60) 8.41 (7.16–9.66) (24.70–45.73) 49.08 (42.45–55.70)
3.62 13.25 9.19 55.71
Neurosurgery (N ¼ 17)
(2.93–4.32) 4.41 (3.93–4.89) (11.02–15.48) 15.56 (14.27–16.84) (7.14–11.24) 12.06 (10.77–13.35) (46.41–65.02) 71.30 (64.87–77.74)
TR-S: Thase and Rush staging; MGH-S: Massachusetts General Hospital staging method; MSM: Maudsley staging method; ATHF: Antidepressant treatment history form; DATA: Dundee antidepressant treatment adequacy score.
234
H. Hazari et al. / Journal of Affective Disorders 150 (2013) 231–236
Table 3 Mean staging method scores between tiers (95% confidence interval).
TR-S MGH-S No. of ‘adequate’ treatments (ATHF) DATA
Primary care (N ¼ 14)
Secondary care (N ¼20)
Tertiary care (N ¼ 67)
0.43 0.54 0.43 3.69
1.90 6.90 4.40 28.97
3.61 12.60 9.52 56.30
(0.06–0.80) (0.02–1.05) (0.06–0.80) (0.53–6.86)
(1.35–2.45) (5.13–8.67) (3.10–5.70) (21.97–35.98)
(3.30–3.93) (11.56–13.64) (8.62–10.42) (51.59–61.01)
TR-S: Thase and Rush staging; MGH-S: Massachusetts General Hospital staging method; MSM: Maudsley staging method; ATHF: Antidepressant treatment history form; DATA: Dundee antidepressant treatment adequacy score.
Table 4 Putative cut-offs for secondary and tertiary care based on MGH-S and ATHF scores.
MGH-S ATHF no. of adequate treatments
Secondary care
Tertiary care
4.00 2.25
9.00 7.00
There is, however, a considerable degree of overlap between the confidence intervals of each of these groups, which is to be expected given the heterogeneity of such groups. This overlap may limit the ability of each staging method to discriminate between levels which are often, to some extent, artificial. Being able to distinguish between primary and secondary care, and between secondary and tertiary care may be more useful in everyday clinical practice. Therefore, these categories were used to compare patient groups. Secondary care patients were constituted from ADC patients and ECT clinic patients (N ¼20), and tertiary care patients were made up of tertiary care referrals, VNS patients, and neurosurgery patients (N ¼ 67). Results are shown below in Table 3. Comparisons on the MGH-S were made between each tier were using t-tests. The difference between each tier (primary/secondary and secondary/tertiary) was statistically significant (Po 0.001) for each measure used Table 4. MGH-S scores for these groups are shown below in Fig. 1.
5. Discussion 5.1. Conclusions Whilst it is challenging to numerically quantify the degree of antidepressant treatment resistance, these data suggest that currently available staging methods may have some capacity to differentiate between clinically-relevant patient groups. By ‘benchmarking’ the levels of treatment received in specific groups, it may be helpful to identify those patients who would benefit from either specific interventions (e.g. ECT) or from specialist clinical review. In Scotland, for example, the standards for integrated care pathways (ICPs) require that specialist assessment should be available for those with treatment-refractory depression (NHS QIS, 2007). Being able to identify those patients who would benefit most would assist services in identifying patients within different parts of the existing care pathway. In addition, such a rating might also highlight those individuals whose previous treatments would prompt the treating clinician to review management in more detail. Whilst the MGH-S and DATA scores would appear to have an equally satisfactory ability to differentiate between different treatment groups (and would appear more sensitive than the TR-S), the relative ease of completion for the MGH-S method would make this a preferable choice for routine clinical practice. Indeed, the advantages of more complex methods of assessing treatment
resistance are less clear when the additional efforts required to calculate them are considered. Whilst ATHF scores (and the derived DATA scores) may have a role in research settings and for defining inclusion criteria for clinical trials, the MGH-S would appear to possess the ability to differentiate between different levels of resistance, as well as providing a continuous variable that is responsive to further treatment. Indeed, we have yet to identify ceiling effects using the MGH-S or DATA scores. The TR-S, when being used to rate low numbers of treatment trials, may be relatively insensitive to potentially therapeutic optimisations of treatment due to its categorical nature. The MSM was not included in this study but further assessment of this tool is planned. We have attempted to provide putative cut-off points at which it would be reasonable to conclude that the patient’s level of resistance is characteristic of the next tier. For each scale, the 25th percentile was considered appropriate since 75% of patients in that tier would be above this score. Such ‘cut-offs’ are shown in Table 4. These figures would suggest that referral to secondary care is occurring after two failed antidepressant trials – this is consistent with the guidelines of NICE (2009) – and referral to tertiary care after seven failed trials (including augmentation strategies). In addition, based on this (small) group of patients, ECT is being delivered to patients who have had four failed antidepressant trials and who have an MGH-S score of 6.0. It is notable that levels of treatment resistance were comparatively low in the primary care sample. However, our findings are consistent with other ratings of treatment in primary care patients with depression. For example, Smith et al. (2011) reported median MGH-S scores of 1.5, reflecting the equivalent of one fullyoptimised antidepressant trial in a population where chronic depression was present in 56%. Even in a US academic psychiatry service, Petersen et al. (2005) found that scores on the MGH-S and TR-S scales were relatively low (the mean 7SD of the entire sample on the MGH-S was 1.6 71.2). 5.2. Limitations This study is the first to compare different methods for describing/staging treatment-refractory depression across such a wide range of illness histories. Treatment histories were based on case note reviews in all cases (thereby avoiding problems due to recall bias) so the likelihood that previous treatments were missed is reduced. Further, the sample is likely to be representative of a real-world clinical population as they were randomly selected from target populations relevant to clinical practice. However, there are weaknesses associated with a study such as this. The numbers are small, and these findings should be considered preliminary. Although attempts were made to ensure that previous treatment trials were not overlooked, the possibility that low levels of treatment in the primary care group reflected insufficient documentation cannot be completely excluded. Further, without being able to corroborate every treatment trial with the patient, it is not possible to always conclude that medication was taken exactly as documented. Clinical correspondence for all groups may, in some cases, overestimate the
H. Hazari et al. / Journal of Affective Disorders 150 (2013) 231–236
235
Fig. 1. Box plot displaying MGH-S Scores for three tiers of service.
robustness of a medication trial. In most cases, assessment of treatment adequacy requires a systematic process of case note review. Assessment of adequacy of previous psychological therapies is an essential process but is complex and would benefit from reliable tools. We agree with Ruhé et al. (2012) that detailed demographics and illness characteristics are preferable in such studies. However, it was not possible to obtain these for all groups in this report; although they are available for the patients in tertiary care and above. We hope to report on these in due course. Using such rating scores to rigidly define the boundaries of clinical services is likely to be premature. It can be seen in Fig. 1 that some patients in the secondary care group had resistance levels comparable with those patients being seen in a tertiary care service. This is likely to reflect relatively resistant patients remaining in secondary care and some patients with lesser degrees of resistance being referred to tertiary care. Social class effects influence waiting times for some surgical interventions (Pell et al., 2000) and may influence referral rates in psychiatry. In the real world, it might be difficult to eliminate crossover; particularly when illness characteristics (e.g. symptom severity, levels of distress, and risk factors such as suicidality) and patient factors (e.g. motivation) will affect referral rates. It is anticipated that larger sample sizes are needed to confirm that these findings are robust across all treatment groups, and further work is required to determine if these ratings can predict outcome. Of course, those with more refractory depression are, by definition, less likely to respond to treatment. A less favourable illness outcome is associated with failure to respond to two or
more drugs and predicting outcome based on indices that already reflect a poor-prognosis group will be challenging. However, it should be possible to determine if response to ECT, for example, varies according to the level of treatment resistance prior to treatment. Currently, research on this prediction is equivocal (Prudic et al., 1990; Kindler et al., 1991). Such a finding would have implications for treatment decisions. Finally, it would be interesting to determine if those patients above a certain threshold who are then referred to more specialist services do better than those who are not referred. Consequently, we would encourage specialist affective disorders services to systematise data collection to aid comparisons of patients and outcomes between services. Our results provide putative support for the use of the MGH-S as a useful research tool.
Role of funding source None.
Conflict of interest HH has received travel and conference attendance from Janssen-Cilag and has had sponsorship for training courses from Wyeth. DC provides clinical management for the Dundee Advanced Interventions Service. He has received consultancy fees from Servier Laboratories and honoraria for lectures from Wyeth and Lilly. He has received travel and conference attendance from Medtronic and Cyberonics Inc. KM has Chaired advisory boards for studies of Deep Brain Stimulation for Obsessive–Compulsive Disorder sponsored by Medtronic. He has received educational grants from Cyberonics Inc & Schering Plough, and he has received research funding from St Jude Medical for a multi-centre clinical trial of deep brain stimulation for depression. He has received travel and accommodation support to attend meetings from Medtronic and St Jude Medical.
236
H. Hazari et al. / Journal of Affective Disorders 150 (2013) 231–236
Acknowledgements We are grateful to primary and secondary care colleagues who allowed us access to clinical records in order to complete the study.
References Barrett, B., Byford, S., Knapp, M., 2005. Evidence of cost-effective treatments for depression: a systematic review. Journal of Affective Disorders 84, 1–13. Berlim, M.T., Turecki, G., 2007. Definition, assessment, and staging of treatmentresistant refractory major depression: a review of current concepts and methods. Canadian Journal of Psychiatry 52, 46–54. Fava, M., 2003. Diagnosis and definition of treatment-resistant depression. Biological Psychiatry 53, 649–659. Fekadu, A., Wooderson, S., Donaldson, C., Markopoulou, K., Masterson, B., Poon, L., Cleare, A.J., 2009a. A multidimensional tool to quantify treatment resistance in depression: the Maudsley staging method. Journal of Clinical Psychiatry 70, 177–184. Fekadu, A., Wooderson, S.C., Markopoulo, K., Donaldson, C., Papadopoulos, A., Cleare, A.J., 2009b. What happens to patients with treatment-resistant depression? A systematic review of medium to long term outcome studies. Journal of Affective Disorders 116, 4–11. Fekadu, A., Wooderson, S.C., Markopoulou, K., Cleare, A.J., 2009c. The Maudsley staging method for treatment-resistant depression: prediction of longer-term outcome and persistence of symptoms. Journal of Clinical Psychiatry 70, 952–957. Judd, L.L., Akiskal, H.S., Maser, J.D., Zeller, P.J., Endicott, J., Coryell, W., Paulus, M.P., Kunovac, J.L., Leon, A.C., Mueller, T.I., Rice, J.A., Keller, M.B., 1998. A prospective 12-year study of subsyndromal and syndromal depressive symptoms in unipolar major depressive disorders. Archives of General Psychiatry 55, 694–700. Kennedy, N., Abbott, R., Paykel, E.S., 2003. Remission and recurrence of depression in the maintenance era: long-term outcome in a Cambridge cohort. Psychological Medicine 33, 827–838. Kindler, S., Shapira, B., Hadjez, J., Abramowitz, M., Brom, D., Lerer, B., 1991. Factors influencing response to bilateral electroconvulsive therapy in major depression. Convulsive Therapy 7, 245–254. McGuffin, P., Farmer, A., Harvey, I., 1991. A polydiagnostic application of operational criteria in studies of psychotic illness. Development and reliability of the OPCRIT system. Archives of General Psychiatry 48, 764–770. National Institute for Health and Clinical Excellence (NICE), 2009. Depression: the Treatment and Management of Depression in Adults. National Collaborating Centre for Mental Health, London. NHS QIS, 2007. Standards for Integrated Care Pathways for Mental Health. NHS Quality Improvement Scotland, Edinburgh.
Oquendo, M.A., Kamali, M., Ellis, S.P., Grunebaum, M.F., Malone, K.M., Brodsky, B.S., Sackeim, H.A., Mann, J.J., 2002. Adequacy of antidepressant treatment after discharge and the occurrence of suicidal acts in major depression: a prospective study. American Journal of Psychiatry 159, 1746–1751. Pell, J.P., Pell, A.C.H., Norrie, J., Ford, I., Cobbe, S.M., Hart, J.T., 2000. Effect of socioeconomic deprivation on waiting time for cardiac surgery: retrospective cohort study. British Medical Journal 320, 15–19. Petersen, T., Papakostas, G.I., Posternak, M.A., Kant, A., Guyker, W.M., Iosifescu, D.V., Yeung, A.S., Nierenberg, A.A., Fava, M., 2005. Empirical testing of two models for staging antidepressant treatment resistance. Journal of Clinical Psychopharmacology 25, 336–341. Prudic, J., Sackeim, H.A., Devanand, D.P., 1990. Medication resistance and clinical response to electroconvulsive therapy. Psychiatry Research 31, 287–296. Ruhé, H.G., van Rooijen, G., Spijker, J., Peeters, F.P.M.L., Schene, A.H., 2012. Staging methods for treatment resistant depression. A systematic review. Journal of Affective Disorders 137, 35–45. Rush, A.J., Trivedi, M.H., Wisniewski, S.R., Nierenberg, A.A., Stewart, J.W., Warden, D., Niederehe, G., Thase, M.E., Lavori, P.W., Lebowitz, B.D., McGrath, P.J., Rosenbaum, J. F., Sackeim, H.A., Kupfer, D.J., Luther, J., Fava, M., 2006. Acute and longer-term outcomes in depressed outpatients requiring one or several treatment steps: a STARnD report. American Journal of Psychiatry 163, 1905–1917. Sackeim, H.A., 2001. The definition and meaning of treatment-resistant depression. Journal of Clinical Psychiatry 62 (Suppl. 16), 10–17. Smith, D.J., Griffiths, E., Kelly, M., Hood, K., Craddock, N., Simpson, S.A., 2011. Unrecognised bipolar disorder in primary care patients with depression. British Journal of Psychiatry 199, 49–56. SPSS, 2007. Version 16.0.2 for Microsoft Windowss. SPSS Inc, Chicago, IL. Stimpson, N., Agrawal, N., Lewis, G., 2002. Randomised controlled trials investigating pharmacological and psychological interventions for treatment-refractory depression. Systematic review. British Journal of Psychiatry 181, 284–294. Thase, M.E., Rush, A.J., 1997. When at first you don't succeed: sequential strategies for antidepressant nonresponders. Journal of Clinical Psychiatry 58 (Suppl 13), 23–29. Thase, M.E., Rush, A.J., Howland, R.H., Kornstein, S.G., Kocsis, J.H., Gelenberg, A.J., Schatzberg, A.F., Koran, L.M., Keller, M.B., Russell, J.M., Hirschfeld, R.M., LaVange, L.M., Klein, D.N., Fawcett, J., Harrison, W., 2002. Double-blind switch study of imipramine or sertraline treatment of antidepressant-resistant chronic depression. Archives of General Psychiatry 59, 233–239. Williams, J., Farmer, A.E., Ackenheil, M., Kaufmann, C.A., McGuffin, P., 1996. A multicentre inter-rater reliability study using the OPCRIT computerized diagnostic system. Psychological Medicine 26, 775–783. World Health Organisation, 1992. International Statistical Classification of Diseases and Related Health Problems, 10th Revision. World Health Organisation, Geneva.