SEARCH

SEARCH BY CITATION

Keywords:

  • algorithm-guided treatment;
  • clinical outcome;
  • lithium augmentation;
  • major depression;
  • treatment as usual

Abstract

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

Aims:  The remission rates for patients with major depressive disorder (MDD) during algorithm-guided treatment (AGT), which consisted of four treatment strategy steps were prospectively compared with treatment as usual (TAU).

Methods:  The remission rates of patients with mild or moderate MDD during AGT (n = 83) were compared with TAU (n = 127).

Results:  The remission rate in the AGT group (60.2%) was approximately 10% greater than that in the TAU group (49.7%). The median number of days to achieve remission in the AGT group (93 days) was half as long as that in the TAU group (191 days). The hazard ratio of remission was 1.5 (95% confidence interval: = 1.2–1.8). A higher rate of lithium augmentation in the AGT group (20.5%) compared to theTAU (4.7%) may have led to the greater remission rate. Most participants who did not achieve remission either during the initial or second treatment steps dropped out from AGT.

Conclusions:  AGT may be superior to TAU for patients with mild or moderate MDD, based on the remission rates achieved. The later treatment steps in the AGT, however, were rarely utilized because participants who did not receive any benefit dropped out early.

PHYSICIANS SHOULD TRY to treat patients with psychiatric disorders logically or on the basis of scientific evidence. Physicians often prescribe medication, however, based on their personal experience rather than evidence-based information. As a consequence there is much individual variation among physicians, and some strategies may be ineffective and therefore prolong symptoms that otherwise might disappear if proven medications were adopted. This is especially true for the treatment of major depressive disorder (MDD) because of the complexity of the symptoms and the clinical course. A possible method for overcoming these problems may be a systematic treatment algorithm with the optimal application, sequencing, and appropriate decision making based on scientific evidence.

Treatment algorithms provide three types of guidance: (i) strategies as to what treatments to use; (ii) tactics on how to implement the treatments; and (iii) treatment steps in a prescribed order to implement the different treatments.1 The critical decision points are defined in the course of treatment, at which time the therapeutic response is to be assessed. On the basis of this assessment, specific treatment revisions are recommended according to the preset if–then rules. Treatment algorithms should determine for how long a single treatment is maintained and how rapidly the treatment plan can be revised.1 Too-frequent decision points bear the risk of not allowing enough time for a specific treatment to show therapeutic benefit. Too-infrequent decision points risk leaving patients on an ineffective treatment for too long. Therefore, the treatment algorithms should set up the optimal intervals between critical decision points based on evidence-based information. But because only a few trials compared next-step treatments in non-responders with initial antidepressant treatment,2–4 the second or third treatment steps are still to be determined. Although actual treatment algorithms remain incomplete, a previous study showed that an algorithm-guided treatment (AGT) for MDD was superior to treatment as usual (TAU).

Some previous studies on AGT of MDD have been conducted.5–7 The Texas Medication Algorithm Project (TMAP) was the first large-scale randomized controlled multicenter study to evaluate treatment algorithms in outpatients with MDD.8 TMAP included seven steps of medication algorithms for the treatment of non-psychotic MDD. Stages 1–3 were monotherapy with an antidepressant. Stages 4–6 were augmentation of an antidepressant with lithium, combination therapy with multiple antidepressants, and electroconvulsive therapy, respectively. The study compared 3-month and 12-month outcomes between AGT and TAU, and found a substantially greater benefit in terms of depressive symptoms, function, and side-effect burden for the algorithm group.

The present study is the first to evaluate AGT in outpatients with MDD in Japan. We created the Saitama Medication Algorithm Project (SMAP) for outpatients with MDD based on the Japanese Psychopharmacology Algorithm Project.9,10 The SMAP algorithm included four treatment steps: the initial three steps were antidepressant monotherapy similar to TMAP. Augmentation therapy with lithium was performed at each step whenever a patient did not achieve remission. We adopted lithium augmentation because lithium has been rarely used in the actual clinical setting even though augmentation with lithium has the strongest research support.11

The standard initial therapy for MDD is monotherapy with an antidepressant, but only 30–50% of patients respond to the first monotherapy.12 Further treatment interventions such as antidepressant augmentation is commonly recommended when patients fail to respond to two or more trials of antidepressant monotherapy. Treatment guidelines most often mention lithium, thyroid supplements, a second ‘add-on’ antidepressant, second-generation antipsychotic medications, and buspirone as potentially helpful augmenting agents.13 Among these augmentation strategies, lithium has the strongest research support. Crossley and Bauer undertook a meta-analysis concerning 10 randomized controlled trials of lithium augmentation of tricyclic antidepressants (TCA) or selective serotonin re-uptake inhibitors (SSRI).14 The authors found that lithium augmentation compared to placebo treatment had a substantially greater response rate with a number needed to treat (NNT) of 5, suggesting that lithium augmentation is the first-choice treatment procedure for patients with MDD who fail to respond to antidepressant monotherapy. Surprisingly, however, lithium has been used rarely in clinical settings.

Recently, Valenstein et al. examined the prevalence and characteristics of antidepressant augmentation in clinical settings.11 Twenty-two percent of patients with depression received an augmentation agent. The most commonly used agents were an add-on second antidepressant (11%) and a second generation antipsychotic (7%). Only 0.5% of the patients received lithium, however, suggesting that lithium augmentation may be a vanishing practice. The treatment outcome of MDD might be more improved if AGT adopts lithium augmentation in a positive manner. To examine the effectiveness of the SMAP algorithm, we compared the clinical results for patients with MDD during AGT, which relied heavily on lithium augmentation, versus TAU.

METHODS

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

Algorithm intervention

Figure 1 shows the SMAP algorithm for patients with mild and moderate MDD without psychotic features, which was created based on the Japanese Psychopharmacology Algorithm Project.9,10 The algorithm consisted of four treatment steps: line 1, monotherapy with SSRI or a selective serotonin norepinephrine re-uptake inhibitor (SNRI); line 2, switching between SSRI and SNRI; line 3, TCA; line 4, a combination of SNRI and TCA. The prescribed dosage was increased when a patient did not achieve remission. If a patient did not reach remission when the dosage of the antidepressant reached the maximum, lithium was added as augmentation therapy in each case.

image

Figure 1. Saitama Medication Algorithm Project (SMAP) algorithm for the treatment of major depressive disorder. Lines 1–3 are monotherapy of an antidepressant. Line 1, selective serotonin re-uptake inhibitor (SSRI) or selective serotonin norepinephrine re-uptake inhibitor (SNRI); line 2, switching between SSRI and SNRI; line 3, tricyclic antidepressant (TCA); line 4, combination of SNRI and TCA. Lithium (Li) augmentation is performed whenever a participant has either a partial or no response.

Download figure to PowerPoint

The SSRI and SNRI available in Japan at the time of the research planning were paroxetine (SSRI), fluvoxamine (SSRI), and milnacipran (SNRI). Participants who entered line 1 were randomly assigned to one of the three SSRI and SNRI. The initial prescribed doses of paroxetine, fluvoxamine, and milnacipran were 10 mg, 50 mg, and 25 mg per day, respectively. Critical decision points were scheduled every 2 weeks. If a participant did not show a satisfactory response to the medication, the dose of each antidepressant was increased to the maximum within 8 weeks. The maximal doses of paroxetine, fluvoxamine, and milnacipran were 40 mg, 150 mg, and 100 mg, respectively. When a participant did not respond to the maximal dose, lithium was added. Lithium augmentation was started at 400 mg per day. If a participant did not show a response, lithium was increased by 200 mg every 2 weeks up to a blood concentration of 0.8 mEq/L. A participant was encouraged to enter line 2 when an antidepressant could not be increased to its maximum due to intolerance or when a participant did not reach remission despite the augmentation of the maximal dose of antidepressant with lithium. If SNRI was used during line 1, the participants were randomly assigned to either paroxetine or fluvoxamine. When a participant entered treatment line 3, imipramine (30–225 mg/day) was administered. Benzodiazepines as anxiolytics or hypnotics could be added but not other psychotropics such as second-generation antipsychotics. Each AGT or TAU physician who actually treated the participant assessed his/her outcome. The outcome measure was whether participants achieved symptom remission using a 5-point (much worse, somewhat worse, just the same, somewhat better, much better) Clinical Global Improvement (CGI) scale.15 Remission was defined as when the CGI score was maintained at 1 (much better) over 4 weeks while a partial response was defined as when a participant did not reach remission but the CGI score was <3.

Participants

The participants were outpatients who first visited hospital between April 2002 and March 2006 and met the DSM-IV criteria for mild or moderate MDD without psychotic features. The exclusion criteria included comorbidity with personality disorders, age under 20 years, and a severe physical condition. Patients who had received antipsychotics, mood stabilizers, anticonvulsants, or two or more types of antidepressants at the initial visit were also excluded. Finally, a total of 240 outpatients were enrolled. Eighty-three and 157 outpatients were assigned to the AGT and TAU groups, respectively. Written informed consent was obtained from participants after a detailed description of the study, which was approved by the Ethics Committee of National Defense Medical College. TAU was performed using the physician's individual discretion based on his/her medical experience and knowledge. Both AGT and TAU physicians had more than 5 years of experiences. At the end of outcome evaluation (October 2006; follow-up period, 6–54 months), the outcomes of all the participants in the AGT group were fixed (remission or dropout), whereas 30 participants in the TAU group continued to be treated. Therefore, 127 out of the 157 participants in the TAU group were included in the present analysis. Table 1 shows the baseline characteristics of the AGT and TAU groups. Age at initial visit, sex ratio, schooling history, marital status, employment status, family size, proportion of subjects living alone, the proportion having concurrent alcohol problems, number of concurrent medical conditions, length suffering from MDD, and the first visit ratio were not different between the two groups.

Table 1.  Baseline subject characteristics
 AGTTAU
(n = 83)(n = 127)
  • Any clinical characteristics do not show significant differences between the groups. Continuous and dichotomous valuables were analyzed using t-test (two-tailed) and χ2 test, respectively.

  • AGT, algorithm-guided treatment; TAU, treatment as usual.

Age (years), mean ± SD49.9 ± 17.148.2 ± 16.9
Female (%)49.449.0
Years in school12.5 ± 2.412.6 ± 2.3
Marital status (%)  
 Married73.563.8
 Divorced1.23.2
 Never married21.726.8
 Separated0.01.6
 Widowed3.64.7
Employment (%)  
 Full time59.048.0
 Part-time9.69.4
 Unemployed31.342.5
Family size (n)1.9 ± 1.22.0 ± 1.3
Living alone (%)9.613.4
Concurrent alcohol problems (%)6.07.9
Concurrent medical condition (n)0.9 ± 1.00.8 ± 1.1
Length of illness (years)3.3 ± 5.84.3 ± 8.8
First visit (%)60.254.3

Statistical analyses

Differences in clinical results between the AGT and TAU groups were analyzed using Kaplan–Meier curves and Cox's proportional hazard model. Moreover, to explore whether an interaction between AGT and lithium augmentation strategy affected outcome, the remission ratios were analyzed using a χ2 test. P < 0.05 was considered statistically significant.

RESULTS

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

Algorithm-guided treatment

In the AGT group the remission, partial or no response, and dropout rates during line 1 were 50.6%, 21.7%, and 27.7%, respectively. These ratios during lines 1 and 2 were 59.0%, 3.6%, and 37.4%, respectively. Finally, 50 out of 83 participants (60.2%) in the AGT group achieved remission while 33 (39.8%) dropped out. Of the dropouts, 15 (18.1%) were unable to tolerate medication. Fourteen participants received lithium augmentation during line 1 while five did during line 2; two of them took lithium during both lines 1 and 2. The dose range of administered lithium was 400–1200 mg/day and the blood concentration ranged from 0.22 to 0.98 mEq/L. Eleven out of 17 participants (70.6%) who received lithium augmentation achieved remission. Three participants (17.6%) dropped out due to intolerance to lithium augmentation. Only three participants entered line 3. No participants entered line 4.

Treatment as usual

Of 127 participants in the TAU group, 82 (64.5%), 17 (13.4%), nine (7.1%), and 19 (15.0%) initially received SSRI, SNRI, TCA or tetracyclic antidepressants, and other antidepressants, respectively. Six participants (4.7%) received lithium augmentation. Finally, 62 participants (49.7%) in the TAU group achieved remission while 65 (51.2%) dropped out. Twelve (9.4%) of the dropouts had intolerance to the medication.

Comparison between AGT and TAU

Figure 2 shows the Kaplan–Meier curves of remission rates comparing the two groups. The remission rate and median number of treatment days to remission in the AGT and TAU groups were 60.2% and 49.7% and 93 days and 191 days, respectively. The hazard ratio of the AGT group compared to the TAU group was 1.49 (95% confidence interval [CI]: 1.21–1.84; χ2 = 13.9, d.f. = 1, P < 0.0002), indicating that AGT was substantially superior to TAU for patients with MDD. The dropout rate due to intolerance (18.1%) in the AGT group was twofold that (9.4%) in the TAU group. Contrary to AGT, TAU physicians could flexibly decide next-step treatment when the ongoing medication could not be continued because of adverse effects. Therefore, the flexibility of TAU might reduce the intolerance rate.

image

Figure 2. Comparison of percentage of patients achieving remission between the algorithm-guided treatment (AGT) and treatment-as-usual (TAU) groups. The remission rate and median number of treatment days to remission in the AGT and TAU groups are 60.2% and 49.7% and 93 days and 191 days, respectively. The hazard ratio of the AGT group compared to the TAU group is 1.49 (95% confidence interval: 1.21–1.84, χ2 = 13.9, d.f. = 1, P < 0.0002).

Download figure to PowerPoint

Lithium augmentation

Because the proportion of the participants who received lithium augmentation in the AGT group (20.5%) was significantly greater than that in the TAU group (4.7%; χ2 = 12.8, d.f. = 1, P < 0.001), differences in outcome between the two groups may be attributed to the differences in the ratio of lithium augmentation. To examine this hypothesis, relationships between treatment group, lithium augmentation, and remission rate were evaluated. The remission rates in the two groups without lithium augmentation were the same (45.8% vs 45.7%), while the rates increased by 14.5% in the AGT group and 3.1% in the TAU group if participants received lithium, indicating that the difference in the remission rates between the groups directly reflected the difference in the number of participants treated with lithium augmentation (χ2 = 12.2, d.f. = 3, P < 0.01). The intolerance rate (17.6%) to lithium augmentation was similar to that (18.1%) in the AGT group as a whole.

DISCUSSION

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

The present results suggest that AGT is superior to TAU for patients with mild or moderate MDD based on their respective remission rates. The remission rate in the AGT group was approximately 10% greater than that in the TAU group. The median number of days to achieve remission in the AGT group was half that in TAU group. The hazard ratio of the remission rate in the AGT group compared to the TAU group was 1.5 (95%CI: 1.2–1.8).

The superiority of AGT in treatment outcome may be explained by two reasons. The first is that AGT requires physicians to decide whether the ongoing medication should be maintained or changed every 2 weeks. Such critical decision points do not exist in TAU. These critical decision points in AGT might reduce the time it takes to achieve remission compared to that for those treated with TAU. Future studies should examine how long TAU physicians maintain the same medication if a patient has a partial response.

The second reason for the superiority of AGT is that the present AGT protocol adopted lithium augmentation in a positive and consistent manner. Lithium currently has the most research support and is well-documented as being the strongest augmentation strategy in refractory depression.16 Interestingly but perhaps expectedly, the remission rate for antidepressant monotherapy without lithium was similar in both the AGT and TAU groups. Therefore, we concluded that the higher rate of lithium augmentation in the AGT group led to their improved remission rate. The maintenance of lithium augmentation, however, is burdensome.

The reasons for the low penetration of lithium augmentation in TAU are unclear but may include concerns about safety, convenience, and tolerability.11 Unlike other augmentation strategies, lithium requires regular laboratory monitoring of medication blood levels, with therapeutic doses sometimes being close to toxicity. Several of lithium's side-effects may appear shortly after treatment initiation. Lithium may also cause renal impairment and thyroid dysfunction. Although the present results again supported the efficacy of lithium, lithium augmentation may become non-existent unless AGT become widely used.

Although the present AGT consisted of four treatment steps, only three out of 83 participants entered the treatment line 3 and no one entered line 4. Whatever further treatment steps such as lines 4 and 5 are prepared for outpatients who have not had a satisfactory response to an initial and one or more subsequent treatments, it is likely that no one will undergo the later treatment steps. We found most outpatients dropped out from AGT when they did not feel a satisfactory response to treatment lines 1 and 2. Although we did not follow up the participants who dropped out, most of them might possibly visit another medical center. Because the therapeutic focus of the present AGT was outpatients with only mild or moderate MDD, we should emphasize that AGT is not applicable for outpatients with severe MDD. Future studies should examine these issues.

The present study had some methodologic limitations: the outcome of the participants was not assessed blindly and rating scales for MDD other than CGI were not used. We believe, however, that the present results suggest the efficacy and efficiency of algorithm-guided treatment for patients with MDD.

ACKNOWLEDGMENTS

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

This study was supported by a grant 17A-5 for nervous and mental disorders from the Ministry of Health, Labour, and Welfare, Japan and by funds from Astellas Pharma, Eli Lilly Japan and Otsuka Pharmaceutical.

REFERENCES

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. ACKNOWLEDGMENTS
  7. REFERENCES
  • 1
    Adli M, Bauer M, Rush AJ. Algorithms and collaborative-care systems for depression: Are they effective and why? A systematic review. Biol. Psychiatry 2006; 59: 10291038.
  • 2
    Gaynes BN, Rush AJ, Trivedi MH, Wisniewski SR, Spencer D, Fava M. The STAR*D study: Treating depression in the real world. Cleve. Clin. J. Med. 2008; 75: 5766.
  • 3
    Papakostas GI, Fava M, Thase ME. Treatment of SSRI-resistant depression: A meta-analysis comparing within- versus across-class switches. Biol. Psychiatry 2008; 63: 699704.
  • 4
    Quitkin FM, McGrath PJ, Stewart JW et al. Remission rates with 3 consecutive antidepressant trials: Effectiveness for depressed outpatients. J. Clin. Psychiatry 2005; 66: 670676.
  • 5
    Crismon ML, Trivedi M, Pigott TA et al. The Texas medication algorithm project: Report of the Texas consensus conference panel on medication treatment of major depressive disorder. J. Clin. Psychiatry 1999; 60: 142156.
  • 6
    Adli M, Berghofer A, Linden M et al. Effectiveness and feasibility of a standardized stepwise drug treatment regimen algorithm for inpatients with depressive disorders: Results of a 2-year observational algorithm study. J. Clin. Psychiatry 2002; 63: 782790.
  • 7
    Bondolfi G, Aubry JM, Golaz J, Gex-Fabry M, Gervasoni N, Bertschy G. A stepwise drug treatment algorithm to obtain complete remission in depression: A Geneva study. Swiss Med. Wkly. 2006; 136: 7885.
  • 8
    Trivedi MH, Rush AJ, Crismon ML et al. Clinical results for patients with major depressive disorder in the Texas Medication Algorithm Project. Arch. Gen. Psychiatry 2004; 61: 669680.
  • 9
    Motohashi N. Algorithm for the Treatment of Mood Disorder (in Japanese). Jiho, Tokyo, 2003.
  • 10
    Nomura S, Sawamura T, Kobayashi N, Yoshino A. Medication algorithm for mood disorders: Present status and future direction in Japan. Int. J. Psychiatry Clin. 2004; 8: 139145.
  • 11
    Valenstein M, McCarthy JF, Austin KL, Greden JF, Young EA, Blow FC. What happened to lithium? Antidepressant augmentation in clinical settings. Am. J. Psychiatry 2006; 163: 12191225.
  • 12
    Nierenberg AA, Amsterdam JD. Treatment-resistant depression: Definition and treatment approaches. J. Clin. Psychiatry 1990; 51 (Suppl.): 3947.
  • 13
    Dording CM. Antidepressant augmentation and combinations. Psychiatr. Clin. North Am. 2000; 23: 743755.
  • 14
    Crossley NA, Bauer M. Acceleration and augmentation of antidepressants with lithium for depressive disorders: Two meta-analyses of randomized, placebo-controlled trials. J. Clin. Psychiatry 2007; 68: 935940.
  • 15
    Guy W. Early Clinical Drug Evaluation (ECDEU) Assessment Manual. National Institute for Mental Health, Rockville, MD, 1976.
  • 16
    Bschor T, Bauer M. Efficacy and mechanisms of action of lithium augmentation in refractory major depression. Curr. Pharm. Des. 2006; 12: 29852992.