Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2013 Mar 4.
Published in final edited form as: Adm Policy Ment Health. 2011 Nov;38(6):486–494. doi: 10.1007/s10488-011-0335-2

Trajectories of Symptom Reduction During Treatment for Behavior Problems in Pediatric Primary-Care Settings

Oliver Lindhiem 1, David J Kolko 1
PMCID: PMC3587333  NIHMSID: NIHMS442453  PMID: 21301952

Abstract

We examined trajectories of symptom reduction during the course of on-site treatment for behavior problems in pediatric primary-care settings for a subset of children from a larger clinical trial. Participants (N = 80) were children (ages 6–11) referred for treatment due to moderate elevations (≥ 75th %ile) on the externalizing subscale of the Pediatric Symptom Checklist (PSC-17). The sample was recruited from six pediatric practices in an urban setting. The treatment (Protocol for an Office-based Nurse-administered Intervention; PONI) was administered on-site by trained nurses over the course of three to six months. Overall, symptom reduction from the first session to last session was moderate (ES = 0.61) and gradual. Roughly two-thirds (72%) of symptom reduction took place over the first eight weeks of treatment. The average trajectory of symptom reduction was characterized by a negative slope that flattened over time, consistent with a quadratic growth model. Initial symptom severity predicted final symptom severity (r = .36, p < .01), even though patients with greater initial symptom severity remained in treatment for a significantly greater number of sessions (r = .12, p < .05). The implications of these findings for enhancing the efficiency and effectiveness of treatment are discussed.

Childhood disruptive behavior problems are common in pediatric primary-care settings where access to specialty mental health services is often limited (Costello, 1989; Lavigne et al., 1998). Even when referrals for appropriate mental health services are available, families often fail to follow through with accessing these services due to numerous obstacles and barriers including lack of financial resources, transportation issues, and stigma associated with mental health services (Kazdin, Holland, & Crowley, 1997; Kazdin & Wassell, 1999). As a result, there are increasing efforts to examine the feasibility, acceptability, and effectiveness of treating mental and behavioral disorders in primary-care settings through collaborations with physicians, nurses, and other providers (e.g., Lavigne, LeBailly, Gouze, Binns, Keller, & Pate, 2010; Turner & Sanders, 2006). The objective of the current study was to examine trajectories of symptom reduction during the course of a psychosocial treatment for disruptive behavior problems delivered in pediatric primary-care settings, which has demonstrated its effectiveness in a randomized trial (reference removed for blind review). Having established the effectiveness of this nurse-administered intervention relative to usual care, the current study is a follow-up examination of session-to-session symptom reduction. Data on the shape and timing of symptom reduction have potential clinical utility for calibrating parent expectations about the likely duration and typical progress that can be expected for a particular treatment (Lindhiem & Kolko, 2010). An understanding of when change takes place during treatment also has implications for determining the optimal number of treatment sessions and consequently improving the efficiency and cost-effectiveness of a given treatment.

Trajectories of Symptom Reduction

There is increasing evidence that symptom reduction during treatment tends to be nonlinear, with most symptom reduction taking place relatively early in treatment (e.g., Hayes, Laurenceau, Feldman, Strauss, & Cardaciotto, 2007; Ilardi & Craighead, 1994; Tang & DeRubeis, 1999). Although many of the early studies of change trajectories were done with treatments for adult depression, recent studies have examined trajectories of symptom reduction for child and adolescent psychosocial treatments for a range of disorders, including anxiety disorders, mood disorders, and disruptive behavior disorders (e.g., Gaynor, Weersing, Kolko, Birmaher, Heo, & Brent, 2003; Warren, Nelson, Mondragon, Baldwin, & Burlingame, 2010). These studies have documented important similarities in change trajectories across patient populations and treatments, such as nonlinearity and early gains. However, important differences are beginning to emerge. For example, a recent study reported symptom reduction to be much more gradual during treatment for childhood behavior disorders relative to cognitive-behavioral treatment for adult depression (Lindhiem & Kolko, 2010). Specifically, trajectories during the course of a psychosocial treatment (primarily parent management training plus cognitive-behavioral therapy) for childhood disruptive behavior disorders (Oppositional Defiant Disorder and Conduct Disorder) were characterized by a gradual decline in symptoms during the initial treatment sessions (roughly two months) followed by a deceleration in the rate of change and flatting of the slope.

Parent Expectations about Treatment Progress

Typical change trajectories are important summaries of clinical data that can be presented to parents in an effort to foster realistic expectations about expected progress during treatment. In those instances for which treatment progress is typically slow, parents who expect quick results might get discouraged and prematurely terminate treatment. As noted above, symptom reduction during treatment for disruptive behavior disorders has been shown to be gradual in at least one study (Lindhiem & Kolko, 2010). However, it has been suggested that trajectories of symptom reduction are likely to differ not only by presenting problems and treatment modality, but also by treatment setting (Warren et al., 2010). As such, it is important to explore typical change trajectories across treatment settings. For example, average trajectories of symptom reduction might differ in pediatric primary-care settings, relative to treatment delivered in university settings or a patient’s home. Even within a given treatment setting, typical change trajectories have been found to differ based on parameters such as initial symptom severity (Howard, Moras, Brill, Martinovich, & Lutz, 1996). For example, patients with greater initial symptom severity might evidence a more rapid reduction in symptoms (steeper slope) relative to patients with lesser initial symptom severity. Conversely, patients with greater initial symptom severity may evidence a flatter slope, indicating slower treatment progress.

Treatment Efficiency and Cost Effectiveness

Trajectories of symptom reduction also have implications for treatment efficiency and cost effectiveness. Even when a given treatment is equally effective across treatment settings, the same outcome might be achieved more rapidly in one setting versus another. For example, a recent study comparing the delivery of a modular treatment for disruptive behavior disorders in two settings (clinic vs. community) reported comparable treatment outcomes even though treatment duration was significantly shorter in the clinic condition compared to the community condition (Kolko, Dorn, Bukstein, Pardini, Holden, & Hart, 2009; Lindhiem & Kolko, 2010). Another study of trajectories of symptom reduction for youth psychotherapy reported a more rapid reduction in symptoms in managed care settings compared to community mental health settings (Warren et al., 2010).

Current Study

The purpose of the current study was to examine typical patterns of reduction in ODD/CD symptoms during the course of a psychosocial treatment delivered in pediatric primary-care settings. Based on our research in other treatment settings (reference removed for blind review), we expected that treatment progress would be gradual relative to other disorders. We also expected that symptom reduction would be consistent with a quadratic growth model, characterized by a negative slope that gradually flattens out during the course of treatment. In addition to these primary hypotheses, several exploratory questions were addressed. First, we examined whether variability in change trajectories was associated with initial symptom severity and treatment duration. Second, we examined whether treatment duration was associated with treatment outcome.

Method

Participants

Participants in the study (N = 80) were from the experimental condition of a randomized controlled trial that compared a protocol for an office-based nurse-administered intervention (PONI) to enhanced usual care (EUC) for the treatment of behavior problems identified in pediatric primary-care settings (reference removed for blind review). Out of 163 participants, 83 were randomly assigned to the experimental condition (PONI) and 80 (96.4%) initiated treatment. Participants who were randomly assigned to the EUC condition (N = 80) were referred to off-site mental health professionals and are not included in the current study. The full sample is described in detail in an earlier paper describing the results of the randomized trial (reference removed for blind review). In summary, approximately two-thirds of participants were male (65%) and the majority were Caucasian (78%). Most parents were married (76%) with at least some college education (80%).

Procedure

Participant recruitment

The study was approved by IRBs at both the University of Pittsburgh Medical Center and the Children’s Hospital of Pittsburgh. Participants were recruited from one of six pediatric primary-care practices in the greater Pittsburgh region that were affiliated with the Children’s Hospital of Pittsburgh. These six practices included 30 pediatricians, all of whom contributed referrals to the study. Referral methods included direct referrals from the child’s pediatrician, parental requests, and response to flyers in clinic waiting areas. Eligibility criteria included child age (5–12 years), scores on externalizing problems subscale of the Pediatric Symptom Checklist (PSC-17; Gardner et al., 1999) at or above the 75th percentile, and parent concerns about the child’s behavior. Exclusion criteria included active psychosis, suicidality with a plan, and concurrent mental health treatment.

Treatment content

The treatment (Protocol for an Office-based Nurse-administered Intervention; PONI) was based on “core components” identified in the disruptive behavior disorder treatment literature (e.g., Chorpita, Daleiden, & Weisz, 2005; Garland, Hawley, Brookman-Frazee, & Hurlburt, 2008) and content from our previous treatment-outcome studies of childhood behavior problems (references removed for blind review). The content of the treatment was organized into three core modules and four supplemental modules. The three core modules were 1) parent management training, 2) child cognitive-behavioral treatment, and 3) family psychoeducation and skills training. Parent management training included teaching parents basic principles such as positive reinforcement (e.g., praise and rewards) and contingency management (e.g., loss of privileges). Child cognitive-behavioral treatment included teaching children relaxation techniques and anger-control strategies. Family psychoeducation and skills training included joint sessions to discuss family rules and practice problem-solving. The four supplemental modules were 1) medication management for ADHD, 2) school/teacher consultation, 3) case and crisis management, and 4) child development expectations/peer enrichment. The three core modules were delivered first to all patients, whereas the four supplemental modules were delivered on an as-needed basis.

Treatment delivery

The treatment was delivered on-site by one of two nurses who were hired and trained as primary clinicians to administer the protocol. Nurses without prior mental health training were selected in order to maximize the generalizability of the study results. The nurse-clinicians each received four months of training in addition to direct clinical supervision for the entire duration of the study. Training included observation, role-plays, and feedback from videotaped clinical interactions. The on-going supervision was provided by a masters-level clinician with experience treating disruptive behavior disorders and the content of the treatment protocol. Supervision was supplemented by consultation with a child and adolescent psychiatrist for medication related questions. Each nurse-clinician delivered the treatment at three of the participating sites. Sessions lasted roughly 90 minutes and were administered every two weeks. Treatment was designed to be delivered in six sessions (phase 1) over three months, although additional sessions (up to 12) could be delivered up to six months to address refractory problems or to promote maintenance of gains. This resulted in an average of nine sessions delivered over the course of three to six months. Overall, 89% of participants completed treatment in six or more sessions.

Data collection

Primary data collection, including standardized assessments and diagnostic interviews, were completed before and after treatment by one of two bachelors-level research staff who were blind to treatment condition. The assessment team received extensive training and clinical supervision. Measures specific to the process and progress of treatment were collected on a session-by-session basis by the nurse-clinicians during the treatment sessions. Additional details describing participant recruitment, treatment content, treatment delivery, and data collection can be found in our earlier paper describing the results of the randomized trial (reference removed for blind review).

Measures

Symptoms of ODD/CD were assessed at every session using the Weekly Report of Behavior Problems measure (Lindhiem & Kolko, 2010), a very brief clinical tool designed to balance sensitivity to change with clinical utility. Because the measure was developed to be administered at every session, the items are restricted to symptoms of ODD and CD in order to limit the burden on clinicians and families. The total score is the sum of seven items, three assessing core ODD symptoms (“defiant/argumentative,” “lost temper,” and “hostility”) and four assessing CD symptoms representing the four primary categories of symptoms of this disorder (“aggressive to people or animals,” “destruction of property,” “deceit: lying, theft,” and “rule violation”). The presence or absence of each symptom during the past week (“YES” or “NO”) was assessed in an interview format with the child’s parent or primary caregiver at the beginning of each session. The WROB has been demonstrated to show sensitivity to change during the course of treatment (Lindhiem & Kolko, 2010). The WROB also demonstrates moderate reliability (Cronbach’s α = .77) and convergent validity with other measures of externalizing problems (Lindhiem & Kolko, 2010).

Data Analysis

Preliminary data analyses were conducted using SPSS version 17.0. Trajectories of symptom reduction were modeled using hierarchical linear modeling (HLM; Raudenbush & Bryk, 2002) with sessions nested within participants. We used full maximum likelihood estimation for all models. Based on the extant literature (e.g., Hayes et al., 2007; Lindhiem & Kolko, 2010), we expected nonlinear change in symptom reduction during the course of treatment. Specifically, we expected a quadratic reduction in symptoms characterized by a negative slope that flattens or decelerates during treatment. The level 1 equation for this model was Yti = π0i + π1i(Sessions) + π2i (Session2) + eti, where Yti is the observed WROB score at session t for participant i. Sessions 1, 2, 3 and 4 were coded “0,” “1,” “2,” and “3,” and so on. This coding scheme allows for a straightforward interpretation of the level-1 intercepts as the symptom level at the initial treatment session.

Results

Descriptive Statistics and Correlations

Descriptive statistics for the level-2 variables are presented in Table 1. Overall, weekly ODD/CD symptoms decreased from the first treatment session (M = 2.86; SD = 1.28) to the last treatment session (M = 2.01; SD = 1.50), F(1, 79) = 23.04, p < .01. Although this symptom reduction appears modest (M = 0.85; SD = 1.58), it represents a medium to large effect size (Cohen’s d = .61). Figure 1 depicts a histogram of change in the number of weekly ODD/CD symptoms reported during the course of treatment for the entire sample (i.e., number reported at first session minus number reported at last session). Overall, 52.5% showed a decrease (change score ≥ 1) in weekly symptoms, 30% showed no change (change score = 0) in weekly symptoms, and 17.5% showed an increase (change score ≤ −1) in symptoms. Participants engaged in 1 to 19 treatment sessions (M = 8.74; SD = 2.75).

Table 1.

Descriptive Statistics for Level-2 Variables

Variables Mean SD Min. Max.
Weekly ODD/CD Symptoms at First Session 2.86 1.28 0 7
Weekly ODD/CD Symptoms at Last Session 2.01 1.50 0 6
Change in Weekly ODD/CD Symptoms from First Session to Last Session1 .85 1.58 −3 5
Number of Treatment Sessions 8.74 2.75 1 19
Open in a new tab
1

Positive numbers indicate a reduction in symptoms.

Figure 1.

Figure 1

Open in a new tab

Change in number of weekly ODD/CD symptoms reported from first session to last session. Positive numbers indicate a reduction in symptoms.

Bivariate correlations between the level-2 variables are presented in Table 2. Weekly ODD/CD symptoms at the first treatment session were significantly associated with weekly ODD/CD symptoms at the last treatment session, r = .36, p < .01. Specifically, patients with high levels of symptoms at the first session also tended to have high levels of symptoms at the last session. Weekly ODD/CD symptoms at the first treatment session were also significantly associated with symptom reduction, r = .47, p < .01, and treatment duration, r = .24, p < .05. Specifically, patients with high scores at the first session tended to show greater symptom reduction and also to remain in treatment for a greater number of sessions. Treatment duration was not associated with overall change in symptoms from the first to last treatment sessions r = −.01, p >.05.

Table 2.

Correlations Between Level-2 Variables

Variables 2 3 4
1. Weekly ODD/CD Symptoms at First Session .36** .47** .24*
2. Weekly ODD/CD Symptoms at Last Session - −.66** .21
3. Change in Weekly ODD/CD Symptoms from First Session to Last Session - −.01
4. Number of Treatment Sessions -
Open in a new tab
*

p < .05.

**

p < .01.

Trajectories of Symptom Reduction

Figure 2 depicts the average trajectory of ODD/CD symptoms during the course of treatment for the full sample. On average, half (52%) of all symptom reduction had occurred by session 3 (4 weeks) and two-thirds (72%) had occurred by session 5 (8 weeks). No additional symptom reduction occurred after session 7 (12 weeks), on average. The trajectory suggested a quadratic decline in symptoms. We tested a quadratic growth model using HLM 6.06 as described above. Parameter estimates for this quadratic growth model are presented in Table 3 (Model 1). The significant linear (β = −0.20; p < .01) and quadratic (β = 0.01; p < .05) terms indicated a negative slope that flattened out over time. In other words, the model indicated symptom reduction that slowed down over the course of treatment.

Figure 2.

Figure 2

Open in a new tab

Average trajectory of weekly ODD/CD symptoms for the full sample (N = 80). The solid line indicates the model-predicted trajectory.

Table 3.

Summary of Quadratic Growth Models of Weekly ODD/CD Symptoms During the Course of Treatment

Intercept
Session (Linear Slope)
Session2 (Quadratic Slope)
B SE t p B SE t p B SE t p
Model 1: Unconditional at level-2
 Level 1 Fixed Effects 2.91 0.12 24.02 0.00 −0.20 0.04 −4.96 0.00 0.01 0.00 2.83 0.01
Model 2: Initial Symptom Severity at level-2
 Level 1 Fixed Effects 0.81 0.17 4.89 0.00 0.23 0.07 3.08 0.00 −0.02 0.01 −2.61 0.01
 Level 2 Fixed Effects 0.73 0.06 13.10 0.00 −0.14 0.02 −6.47 0.00 0.01 0.00 4.17 0.00
Model 3: Final Symptom Severity at level-2
 Level 1 Fixed Effects 2.54 0.21 12.30 0.00 −0.26 0.07 −3.54 0.00 0.01 0.01 0.64 0.53
 Level 2 Fixed Effects 0.17 0.09 1.93 0.06 0.05 0.03 1.44 0.15 0.00 0.00 0.46 0.64
Model 4: Treatment Duration at level-2
 Level 1 Fixed Effects 2.33 0.32 7.35 0.00 −0.17 0.11 −1.50 0.14 0.00 0.01 0.03 0.98
 Level 2 Fixed Effects 0.07 0.04 1.61 0.11 −0.00 0.01 −0.12 0.90 0.00 0.00 1.12 0.27
Open in a new tab

Next, we examined whether initial symptom severity accounted for individual differences in trajectories of symptom reduction. To test this, initial symptom severity was added to the growth model as a level-2 predictor. Parameter estimates for this model are presented in Table 3 (Model 2). Patients who started treatment with relatively many symptoms had a greater linear decrease in symptoms (β = −0.14; p < .01). However, they also had a more positive quadratic term (β = 0.01; p < .01), indicating a greater deceleration in symptom reduction during the course of treatment. To visualize the results in Figure 3, we used a median-split to divide the sample into patients who started treatment with relatively many weekly ODD/CD symptoms (3–7) and those who started treatment with relatively few weekly ODD/CD symptoms (0–2).

Figure 3.

Figure 3

Open in a new tab

Trajectories of weekly ODD/CD symptoms for patients who started treatment with 3–7 “high” weekly symptoms (N = 54) and 0–2 “low” weekly symptoms (N = 26). Solid lines indicate the model-predicted trajectories.

We also examined whether final (last session) symptom severity accounted for individual differences in trajectories of symptom reduction. To test this, final session symptom severity was added to the initial quadratic growth model as a level-2 predictor. Parameter estimates for this model are presented in Table 3 (Model 3). Final symptom severity did not predict the linear decrease in symptom reduction (β = 0.05; p > .05) nor the quadratic term (β = 0.00; p > .05). To visualize the results in Figure 4, we again used a median-split to divide the sample into patients who ended treatment with relatively many weekly ODD/CD symptoms (3–7) and those who ended treatment with relatively few weekly ODD/CD symptoms (0–2).

Figure 4.

Figure 4

Open in a new tab

Trajectories of weekly ODD/CD symptoms for patients who ended treatment with 3–7 “high” weekly symptoms (N = 33) and 0–2 “low” weekly symptoms (N = 47). Solid lines indicate the model-predicted trajectories.

Finally, we examined whether treatment duration (“dose”) accounted for individual differences in trajectories of symptom reduction. To test this, the total number of treatment sessions was added to the initial quadratic growth model as a level-2 predictor. Parameter estimates for this model are presented in Table 3 (Model 4). Treatment duration was not associated with the linear decrease in symptom reduction (β = −0.00; p > .05) nor the quadratic term (β = 0.00; p > .05). To visualize the results in Figure 5, we again used a median-split to divide the sample into patients with relatively long treatment duration (9 – 19 sessions) and those with relatively brief treatment duration (1 – 8 sessions).

Figure 5.

Figure 5

Open in a new tab

Trajectories of weekly ODD/CD symptoms for patients who attended nine or more “long” treatment sessions (N = 46) and eight or fewer “brief” treatment sessions (N = 34). Solid lines indicate the model-predicted trajectories.

Discussion

Magnitude of Symptom Reduction

Overall, our results indicated moderate symptom reduction during the course of treatment delivered by nurses in primary-care settings (Cohen’s d = .61), consistent with the treatment outcome data for the current sample (reference removed for blind review). However, it is important to note that there was significant variability within the overall sample. In absolute terms, 30% of the sample did not show any decrease in symptom reduction from the first to last treatment session, and 17.5% showed an increase in symptoms. These percentages are consistent with those reported in the extant literature on youth psychotherapy. In a recent study of youth psychotherapy in managed care settings and community settings, for example, 14–24% of patients showed an increase in symptoms and an additional 31–32% showed no reliable change in symptoms (Warren et al., 2010).

Shape and Timing of Symptom Reduction

Shape

The average trajectory of symptom reduction was characterized by a negative slope that flattened over time. This deceleration in symptom reduction was consistent with a quadratic growth model, and replicates change trajectories for the treatment of disruptive behavior disorders in other samples. For example, a recent study of change in weekly ODD/CD symptoms during the course of a modular treatment for children with disruptive behavior disorders revealed the same initial decrease with a gradual deceleration (Lindhiem & Kolko, 2010). Similar change trajectories have also been described for youth psychotherapy in general (e.g., Cannon et al., 2010, Warren et al., 2010).

Timing

Roughly two-thirds (72%) of symptom reduction took place by session 5 of treatment (8 weeks). This finding is also consistent with a recent study on change trajectories during a psychosocial treatment for disruptive behavior disorders (DBDs), which reported approximately two-thirds of symptom reduction taking place during the first two months of treatment (Lindhiem & Kolko, 2010). In the current study, no additional change in symptom reduction occurred after session 7 (12 weeks). In addition, treatment duration was unrelated to overall symptom reduction or outcome (i.e. symptom level at the final treatment session). This absence of a dose-response effect is consistent with the child and adolescent psychotherapy literature (e.g., Bickman, Andrade, & Lambert, 2002; Salzer, Bickman, & Lambert, 1999).

Individual Differences in Symptom Reduction

In the current study, we were also interested in examining individual differences in trajectories of symptom reduction. We examined associations between change trajectories and initial symptom severity, final symptom severity, and treatment duration. Our results suggested that initial symptom severity was associated with a steeper initial decline in symptom reduction but a greater deceleration in symptom reduction. Overall, patients who began treatment with relatively high levels of symptoms also ended treatment with relatively high levels of symptoms. In other words, patients who started treatment with greater symptom severity did not “catch up” to their counterparts with lesser symptom severity, despite the fact that patients with greater symptom severity tended to remain in treatment for a greater number of treatment sessions.

Limitations and Future Directions

One limitation of this study is the relatively restricted range of our primary measure of symptom reduction. However, the WROB has sufficient variability to model change in symptoms during the course of treatment as evidenced in the current study and prior research (Lindhiem & Kolko, 2010). Although this restricted range might give the impression that many patients were relatively asymptomatic at baseline, these symptoms represent major criteria for ODD/CD that were present in the past week out of a total of seven general symptom areas surveyed. This is a short interval relative to DSM-IV diagnostic criteria. To put this in context, only four criteria must be present during the past six months to meet diagnostic criteria for Oppositional Defiant Disorder. Only three symptoms (albeit more serious ones) must be present during the past 12 months to meet diagnostic criteria for Conduct Disorder. For the same reason, a change in even one weekly symptom from pre to post treatment represents a substantial, and potentially clinically meaningful, change.

A limitation of examining the effects of initial symptom severity and treatment duration on change trajectories in the current study is the covariation among these variables. In particular, patients with greater initial symptom severity tended to stay in treatment for more sessions compared to patients with fewer initial symptoms. Our findings pertaining to the associations between initial symptom severity, treatment duration, and change trajectories must therefore be interpreted as correlational. Another limitation of the study is the absence of session-by-session symptom data for the EUC or “usual-care” condition. Because treatment in the EUC condition was delivered by non-study (community) clinicians, this data could not be collected without altering the nature of the “usual-care” condition. Without trajectory data for a control condition, the results in this study must be viewed as descriptive rather than inferential. In other words, an unknown proportion of the symptom reduction might be due to a placebo effect or regression-to-the-mean rather than the treatment per se. However, given that that the effectiveness of the current treatment (PONI) has been demonstrated in the context of a randomized trial, neither of these possibilities are likely to account for the findings in the current study. Finally, it is possible that weekly symptom counts were influenced by rater or reporter bias. For example, it is possible that parents minimized or exaggerated symptoms in an attempt to shorten or extend the duration of treatment. Although any such bias would have likely resulted in “noise” or “error”, it is always possible that a systematic bias was present across the study.

In terms of future directions, the current study was limited to an examination of only a few variables that might influence trajectories of symptom reduction. Specifically, we focused on variables that, by definition, influence trajectories, namely start point (i.e., initial severity), end point (i.e., final severity), and length (i.e., treatment duration). There are certainly other variables that might influence the shape and timing of symptom reduction, including clinical/patient characteristics, provider characteristics, and the choice of clinical content and supplemental interventions. These variable, and others, are worth examining in future studies. Further development and refinement of measures to monitor treatment progress will also likely lead to improved clinical decision making during the course of treatment.

Summary and Conclusions

Few studies have examined session-by-session symptom reduction during child and adolescent psychosocial treatments, particularly for children with disruptive behavior disorders. This is the only study we are aware of that has examined session-by-session symptom reduction in primary-care settings for this population. Consistent with our prior research findings on treatment course for disruptive behavior disorders in outpatient and community settings (reference removed for blind review), symptom reduction was found to be relatively gradual during the first two months of treatment. Roughly two-thirds of symptom reduction occurred after eight weeks in treatment. Symptom severity at the final treatment session was associated with initial symptom severity but not treatment duration. Initial symptom severity predicted final symptom severity despite the fact that patients with greater initial symptom severity remained in treatment for a significantly greater number of sessions.

Implications for Research, Policy, and Practice

Several results from our study have implications for parents’ expectations about treatment progress in pediatric primary-care settings. First, symptom reduction was found to be gradual over the first eight weeks of treatment. Parents who expect a “quick-fix” to their child’s behavior problems are perhaps more likely to become discouraged by treatment progress and terminate before they have achieved an optimal response. The data from this study suggest that parents should be encouraged to “stick with it” for approximately 2–3 months in order to achieve maximum benefit. Another result that has implications for parents’ expectations is that patients with relatively more symptoms at the beginning of treatment continued to have relatively more symptoms at the end of treatment. This is consistent with meta-analytic findings on the role of initial symptom severity in predicting treatment outcome for behavior problems (Reyno & McGrath, 2006). Initial symptom severity is a robust predictor of treatment outcome.

Our results also suggest that there is little additional symptom reduction after 12 weeks (approximately three months) for this particular psychosocial treatment delivered in pediatric primary-care settings. This finding has implications for the efficiency and cost-effectiveness of treating behavior problems in pediatric primary-care settings. For patients who continue to evidence significant levels of symptoms, it is perhaps an inefficient use of resources to continue to provide the same treatment past three months. Rather, these families might be better served by a referral to a specialty mental health clinic. Alternatively, it is possible that additional treatment development or supplemental content might lead to continued reduction in symptoms beyond the three month period.

Additional research is also needed to determine what evidence-based strategies might be suggested to enhance or accelerate the level of symptom reduction during the first 2–3 months of treatment. Finally, it should also be noted that although we speculate about the possibility, the current study did not directly test whether providing families with typical treatment course data indeed reduces premature drop-out. We would expect that parents with realistic expectations about treatment progress are more likely to complete treatment. However, this remains to be tested empirically in a future study.

Acknowledgments

This study was supported by a grant to the second author from the National Institute of Mental Health (MH63272) and by a T32 Fellowship to the first author. The authors acknowledge the research and clinical staff of the (Services for Kids in Primary-Care) SKIP program.

References

  1. Bickman L, Andrade AR, Lambert EW. Dose response in child and adolescent mental health services. Mental Health Services Research. 2002;4:57–60. doi: 10.1023/a:1015210332175. [DOI] [PubMed] [Google Scholar]
  2. Cannon JAN, Warren JS, Nelson PL, Burlingame GM. Change trajectories for the youth outcome questionnaire self-report: Identifying youth at risk for treatment failure. Journal of Clinical Child & Adolescent Psychology. 2010;39(3):289–301. doi: 10.1080/15374411003691727. [DOI] [PubMed] [Google Scholar]
  3. Chorpita BF, Daleiden EL, Weisz JR. Identifying and selecting common elements of evidence based interventions: A distillation and matching model. Mental Health Services Research. 2005;7(1):5–20. doi: 10.1007/s11020-005-1962-6. [DOI] [PubMed] [Google Scholar]
  4. Costello EJ. Child psychiatric disorders and their correlates: A primary care pediatric sample. Journal of the American Academy of Child & Adolescent Psychiatry. 1989;28(6):851–855. doi: 10.1097/00004583-198911000-00007. [DOI] [PubMed] [Google Scholar]
  5. Gardner W, Murphy JM, Childs G, Kelleher K, Pagano ME, Jellinek MS, Chiapetta L. The PSC-17: A brief pediatric symptom checklist with psychosocial problem subscales. A report from PROS and ASPN. Ambulatory Child Health. 1999;5:225–236. [Google Scholar]
  6. Garland AF, Hawley KM, Brookman-Frazee L, Hurlburt MS. Identifying common elements of evidence-based psychosocial treatments for children’s disruptive behavior problems. Journal of the American Academy of Child and Adolescent Psychiatry. 2008;47(5):505–514. doi: 10.1097/CHI.0b013e31816765c2. [DOI] [PubMed] [Google Scholar]
  7. Gaynor ST, Weersing VR, Kolko DJ, Birmaher B, Heo J, Brent DA. The prevalence and impact of large sudden improvements during adolescent therapy for depression: A comparison across cognitive-behavioral, family, and supportive therapy. Journal of Consulting and Clinical Psychology. 2003;71(2):386–393. doi: 10.1037/0022-006X.71.2.386. [DOI] [PubMed] [Google Scholar]
  8. Hayes AM, Laurenceau J, Feldman G, Strauss JL, Cardaciotto L. Change is not always linear: The study of nonlinear and discontinuous patterns of change in psychotherapy. Clinical Psychology Review. 2007;27(6):715–723. doi: 10.1016/j.cpr.2007.01.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Howard KI, Moras K, Brill PL, Martinovich Z, Lutz W. Evaluation of psychotherapy: Efficacy, effectiveness, and patient progress. American Psychologist. 1996;51(10):1059–1064. doi: 10.1037/0003-066X.51.10.1059. [DOI] [PubMed] [Google Scholar]
  10. Ilardi SS, Craighead WE. The role of nonspecific factors in cognitive behavior therapy for depression. Clinical Psychology: Science and Practice. 1994;1(2):138–156. [Google Scholar]
  11. Kazdin AE, Holland L, Crowley M. Family experience of barriers to treatment and premature termination from child therapy. Journal of Consulting and Clinical Psychology. 1997;65(3):453–463. doi: 10.1037//0022-006x.65.3.453. [DOI] [PubMed] [Google Scholar]
  12. Kazdin AE, Wassell G. Barriers to treatment participation and therapeutic change among children referred for conduct disorder. Journal of Consulting and Clinical Psychology. 1999;28(2):160–172. doi: 10.1207/s15374424jccp2802_4. [DOI] [PubMed] [Google Scholar]
  13. Kolko DJ. Multimodal partial-day treatment of child antisocial behavior: Service description and multi-level program evaluation. Continuum: Developments in Ambulatory Mental Health Care. 1995;2:3–24. [Google Scholar]
  14. Kolko DJ, Campo JV, Kelleher K, Cheng Y. Improving access to care and clinical outcome for pediatric behavioral problems: A randomized trial of a nurse-administered intervention in primary care. Journal of Developmental and Behavioral Pediatrics. 2010;31(5):393–404. doi: 10.1097/DBP.0b013e3181dff307. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Kolko DJ, Dorn LD, Bukstein OG, Pardini D, Holden EA, Hart J. Community vs. clinic-based modular treatment of children with early-onset ODD or CD: A clinical trial with 3-year follow-up. Journal of Abnormal Child Psychology. 2009;37:591–609. doi: 10.1007/s10802-009-9303-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Lambert MJ, Finch AE. The Outcome Questionnaire. In: Maruish ME, editor. The use of psychological testing for treatment planning and outcome assessment. 2. Mahwah, NJ: Earlbaum; 1999. pp. 831–870. [Google Scholar]
  17. Lambert MJ, Hansen NB, Finch AE. Patient-focused research: Using patient outcome data to enhance treatment effects. Journal of Consulting and Clinical Psychology. 2001;69(2):159–172. doi: 10.1037//0022-006X.69.2.159. [DOI] [PubMed] [Google Scholar]
  18. Lambert MJ, Whipple JL, Hawkins EJ, Vermeersch DA, Nielsen SL, Smart DW. Is it time for clinicians to routinely track patient outcome? A meta-analysis. Clinical Psychology: Science and Practice. 2003;10(3):288–301. doi: 10.1093/clipsy/bpg025. [DOI] [Google Scholar]
  19. Lavigne JV, Arend R, Rosenbaum D, Binns HJ, Christoffel KK, Burns A, Smith A. Mental health service use among young children receiving pediatric primary care. Journal of the American Academy of Child & Adolescent Psychiatry. 1998;37(11):1175–1183. doi: 10.1097/00004583-199811000-00017. [DOI] [PubMed] [Google Scholar]
  20. Lavigne JV, LeBailly SA, Gouze KR, Binns HJ, Keller J, Pate L. Predictors and correlates of completing behavioral parent training for the treatment of oppositional defiant disorder in pediatric primary care. Behavior Therapy. 2010;41(2):198–211. doi: 10.1016/j.beth.2009.02.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Lindhiem O, Kolko DJ. Trajectories of symptom reduction and engagement during treatment for childhood behavior disorders: Differences across settings. Journal of Abnormal Child Psychology. 2010;38(7):995–1005. doi: 10.1007/s10802-010-9416-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Raudenbush SW, Bryk AS. Hierarchical linear models: Applications and data analysis methods. 2. Thousand Oaks, CA: Sage; 2002. [Google Scholar]
  23. Reyno SM, McGrath PJ. Predictors of parent training efficacy for child externalizing behavior problems – a meta-analytic review. Journal of Child Psychology and Psychiatry. 2006;47(1):99–111. doi: 10.1111/j.1469-7610.2005.01544.x. [DOI] [PubMed] [Google Scholar]
  24. Salzer MS, Bickman L, Lambert EW. Dose-effect relationship in children’s psychotherapy services. Journal of Consulting and Clinical Psychology. 1999;67:228–238. doi: 10.1037//0022-006x.67.2.228. [DOI] [PubMed] [Google Scholar]
  25. Tang TZ, DeRubeis RJ. Sudden gains and critical sessions in cognitive behavioral therapy for depression. Journal of Consulting and Clinical Psychology. 1999;67(6):894–904. doi: 10.1037/0022-006X.67.6.894. [DOI] [PubMed] [Google Scholar]
  26. Turner KMT, Sanders MR. Help when it’s needed first: A controlled evaluation of brief, preventive behavioral family intervention in a primary care setting. Behavior Therapy. 2006;37(2):131–142. doi: 10.1016/j.beth.2005.05.004. [DOI] [PubMed] [Google Scholar]
  27. Warren JS, Nelson PL, Mondragon SA, Baldwin SA, Burlingame GM. Youth psychotherapy change trajectories and outcomes in usual care: Community mental health versus managed care settings. Journal of Consulting and Clinical Psychology. 2010;78(2):144–155. doi: 10.1037/a0018544. [DOI] [PubMed] [Google Scholar]

RESOURCES