Correlates of Treatment Outcome for Patients With Mood Disorders

In my last blog post, I described end-of treatment outcomes for patients with mood disorders. Here, I will elaborate on factors that were correlated with treatment outcome for these patients. Please be advised that these results are specific to my practice and my patients, and should not necessarily be generalized to other clinicians or populations.

Diagnosis
Type of mood disorder diagnosis was strongly correlated with treatment outcome. Patients with Depressive Disorder NOS (DDNOS; n = 3) fared best. Two-thirds of these patients (n = 2) completed treatment and achieved full remission. The remaining one-third (n = 1) attended treatment for two months and made significant progress before quitting prematurely.

Patients with Mood Disorder NOS (MDNOS; n = 4) had varying outcomes. Half of them (n = 2) completed treatment; of these, one achieved full remission and the other made significant progress. Of the remaining MDNOS patients, one quit prematurely after making some progress and the other regressed and was referred to a higher level of care.

Both of the patients in this sample with Dysthymic Disorder also had a diagnosis of Major Depressive Disorder (MDD). For statistical purposes, I am subsuming these individuals under the category of MDD, as that is the more severe of the two diagnoses. Patients with Major Depressive Disorder (MDD; n = 11) had varying outcomes. Eighteen percent (n = 2) completed treatment and achieved full remission; 9% (n = 1) quit treatment after making significant progress; 45% (n = 5) either quit treatment or moved away after making some progress; and 27% (n = 3) were referred to other providers.

The presence of both dysthymic disorder and MDD (n = 2) did not have a consistent impact on treatment outcome – one of these individuals completed treatment and achieved full remission, while the other regressed and was referred to a higher level of care.

None of the patients with bipolar disorder (n = 3) completed treatment or achieved full remission. However, one of the patients with bipolar disorder remained in treatment for over a year and made significant progress, but was eventually referred to another treatment provider to address other treatment needs.

Comorbidity
Presence of a co-morbid diagnosis was a negative prognostic factor. Only 18% (n = 2) of the patients with comorbid disorders completed treatment, and only 9% (n = 1) achieved full remission, while the remaining 9% (n = 1) made significant progress. In contrast, 40% (n = 4) of patients without a comorbid diagnosis completed treatment, and all of these patients achieved full remission.

Medication
Patients who did not take psychiatric medication during treatment with me were more likely to make progress in treatment and more likely to achieve remission than those who took psychiatric medication. All of the patients who did not take medication (n = 8) made at least some progress in treatment. In contrast, 30% of those who took medication either made no progress (n = 2) or regressed (n = 2) during treatment. Half of the patients who did not take medication (n = 4) achieved remission by the end of treatment, compared to 8% (n = 1) of those who took medication.

My hypothesis is that patients who took medication had poorer outcomes not because their medication didn’t work or made them worse, but rather because taking medication is confounded with severity of mood disorder and with comorbidity. In other words, those with more severe mood disorders and/or comorbid conditions were more likely to require medication in order to function, whereas those with mild to moderate mood disorders were less likely to need medication.

Age and Family Involvement
Younger age was a significant predictor of treatment completion. Patients who completed treatment ranged in age from 16 – 29, with an average age of 19.7. Those who did not complete treatment ranged in age from 12 – 59, with an average age of 28.6.

Being under 18 years old was a protective factor. None of the adolescents under 18 quit treatment prematurely. Of the 6 adolescents in this sample, 50% (n = 3) completed treatment and attained full remission, 33% (n = 2) were referred to other treatment settings after a year or more of treatment with me due to other needs, and the remaining 16% (n = 1) engaged in treatment as a 17-year-old and made some progress, but quit shortly after his 18th birthday. I suspect that the high level of parental involvement that I require for adolescent patients is the primary reason why they are likely to remain in treatment and have positive outcomes. In addition, younger patients tend to have shorter duration of illness compared to adult patients, and early intervention is also predictive of positive outcome.

For patients over 18, family involvement often included spouses and significant others instead of parents, based on the patient’s living situation and relationship status. Sixty percent (n = 9) of the patients over 18 in this sample had no family involvement whatsoever, 27% (n = 4) had a low level of family involvement, and 13% (n = 2) had a moderate level of family involvement. None of the patients over 18 had a high level of family involvement.

Importantly, amongst the adults in this sample, family involvement was not necessary in order to make progress in treatment or attain remission. None of the adults who attained full remission (n = 2) had family members involved in their treatment. Likewise, 67% (n = 2) of the adults who made significant progress in treatment had a low level of family involvement and 33% (n = 1) had no family involvement at all.

None of the patients over age 30 (n = 5) completed treatment. I hypothesize that there are several reasons for this: 1.) Older patients had a longer duration of illness, which means that their illness was more entrenched and more difficult to treat; 2.) The majority of these patients (80%; n = 4) had no family support at all; 3.) All of these older patients paid reduced rates for my services, which indicates that they were of lower socio-economic status, and which may suggest that they value my services less than those who pay full price; 4.) I saw all of these patients during my post-doctoral year, when I was less experienced and not yet fully licensed. Therefore, it is reasonable to assume that the quality of my services was somewhat lower back then than it is now that I am fully licensed and more experienced; and 5.) I prefer working with child and adolescent patients and find that I tend to be more effective with them. For these reasons, particularly #5, I no longer treat adult patients beyond their mid-twenties.

Gender
Average duration of treatment was comparable for males and females (6.8 months vs 7.4 months, respectively). Rates of remission were similar between males and females. Twenty percent (n = 1) of males completed treatment and achieved remission, compared with 25% (n = 4) of females. However, amongst those who did not achieve remission, females were more likely to make significant progress than males. Twenty-five percent (n = 4) of females made significant progress, compared with none of the males.

Hospitalization
Hospitalization during treatment with me predicted regression in treatment. Of the patients who were hospitalized during treatment with me, 67% (n = 2) had regressed by the end of their treatment with me. In contrast, none of the patients who were not hospitalized during treatment had regressed as of their final session with me.

I hypothesize that patients who remained out of the hospital during treatment with me were responding well to treatment, which would explain why they were more likely to achieve remission or at least make significant progress. In contrast, being hospitalized during treatment may have been an indication that the patient was not responding well to treatment, which explains why they were likely regressed as of their last session with me.

Surprisingly, history of hospitalization before entering treatment with me was not related to outcome. Of the 8 patients who had been hospitalized previously, 25% (n = 2) achieved remission, 25% (n = 2) made significant progress, 38% (n = 3) made some progress, and only 12% (n = 1) regressed in treatment with me. These percentages are comparable to those of the 13 patients who had not been hospitalized before beginning treatment with me: 23% (n = 3) achieved remission, 15% (n = 2) made significant progress, 38% (n = 5) made some progress, 15% made no progress, and 8% (n = 1) regressed.

I would have suspected that history of hospitalization would be confounded with illness severity, and therefore would predict poor treatment outcome. However, it is possible that the individuals who had been hospitalized before entering treatment with me had not been given appropriate outpatient treatment, hence the escalation of illness need for hospitalization. Perhaps these patients were able to benefit from their treatment with me, and it is possible that, in some cases, hospitalization could have been avoided if they had received effective outpatient treatment sooner.

Attendance at Therapy Sessions
Attendance at therapy sessions was correlated with treatment retention as well as treatment outcome. Sixty-seven percent (n = 4) of those who completed treatment attended all of their appointments, while the remaining 33% (n = 2) had only one missed appointment. Eighty percent of patients who attained full remission (n = 4) had perfect attendance in therapy; the remaining 20% (n = 1) missed only one appointment. None of the patients who missed more than one appointment completed treatment or achieved full remission.

I suspect this relates to the old adage: “You get out of it what you put into it.” A patient who frequently misses appointments probably places little value on their mental health, which may explain why they tended to drop out of treatment prematurely. Those who were diligent about their treatment prioritized their mental health and worked hard in therapy. Those who were conscientious and responsible about attendance were, most likely, conscientious and responsible about completing therapy homework, taking their medication consistently, and making positive life choices in general. Thus, it is natural that these individuals had better treatment outcomes.

Fee for Services
Payment of full fee for services was a predictor of treatment completion and achieving full remission. Sixty-seven percent of patients (n = 2) who paid my full rate completed treatment, compared with only 22% of patients (n = 4) who paid a reduced rate. Given that so few patients in this sample paid my full rate (n = 3), it is difficult to draw any conclusions about how these individuals differ from those who pay a lower rate.

However, in my practice, the correlation between payment of full services and better treatment outcome has been consistent across diagnoses, regardless of the number of clients in the sample. As I have noted in recent posts, individuals who pay lower fees are, by definition, of lower socio-economic status (SES). These individuals may have more financial stressors than those of higher SES, and may have other life obstacles in general (limited access to healthcare, transportation difficulties, unemployment, single parent families) which interfered with their ability to progress in treatment.

In addition, those who pay reduced rates were much more likely than those who pay full rate to no-show for sessions or to cancel at the last minute. This finding suggests to me that, on average, individuals who paid reduced fees placed less personal value on their sessions, had less respect for my time, did not prioritize their recovery, and/or were generally irresponsible, compared with individuals who paid full rate.