10/21/15 Journal Club

Presenters: Adam Pendleton (PGY3) and Madison Johnson (PGY2)

Special Presenter: Dr. Frank Harrell

Course Director: Dr. Stephan Heckers


Prescription Practices in the Treatment of First-Episode Schizophrenia Spectrum Disorders: Data from the National RAISE-ETP Study; Am J Psychiatry 172:3, March 2015

1.a. What motivated this study? Different treatment guidelines have been utilized by prescribing clinicians for individuals with schizophrenia and schizophrenia-spectrum disorders (schizophreniform, schizoaffective, psychosis NOS, and brief psychotic disorder) based upon whether they are experiencing a first episode of psychosis or have experienced multiple episodes of psychosis. What has not previously been studied is whether community clinicians, who, for many patients, will be the primary provider of mental health care, change their prescribing practices for first-episode patients. This article evaluates both the prescription patterns of community providers, as well as examines what factors play a role in the choice of treatment.

1.b. Why do we care? As schizophrenia is a chronic condition, with most patients experiencing multiple episodes, the initial treatment is important in developing a safe, effective, and sustainable regimen for the patient. These treatment choices have been guided by prior studies (see “Additional Resources” – Section 2). However, whether the medications chosen by community providers follow these guidelines is unknown. By evaluating how and why community clinicians prescribe, we can provide improved patient care. As trainees on the inpatient unit, evidence from this article can assist with the transition to community-based care by explaining treatment choice. Additionally, by using a Bayesian perspective for the correlate analyses, underlying factors that may influence provider’s treatment choice can be identified and corrected.

2.a. Did the authors explicitly state the research question? The authors clearly posed two questions in the introduction of the paper: “What are the medication treatments currently used in community settings?” and, “What factors are associated with choice of medication strategy?” The authors did not include the specific patient population or associated factors in their stated question, but are discussed in the introduction and method sections.

2.b. What is the specific research question that the authors are trying to answer? Among patients with first-episode psychosis (P), are the prescribing practices of community clinicians who choose treatment based on provider choice (I) different than the treatment choices made by employing clinical guidelines (C)? If there is a difference in prescribing practice, what patient factors (region, gender, age, ethnicity, diagnosis, insurance status, and/or comorbid depression and anxiety) (O) are associated with medication choice by community within the first 6 months of treatment (T)?

3.a. What study design was chosen? This article employed a retrospective cohort study to answer the research questions by analyzing the prescription data of 404 first-episode patients at the time of entry to the study. (Further studies randomized the patients into two cohorts – a group whose treatment was guided by clinician choice alone, and another group whose treatment was guided by the NAVIGATE program (“Additional Resources” – Section 2).

3.b. Who was studied? The design prioritized enhancing generalizability of findings to community settings. Inclusion/exclusion criteria were chosen to allow broad inclusion of different patient subgroups. Inclusion criteria were: age 15 to 40 years; diagnosis of schizophrenia, schizophreniform disorder, schizoaffective disorder, psychosis NOS, or brief psychotic disorder; beginning first treatment for psychosis (defined as having taken antipsychotic medications cumulatively for 6 months or less) and ability to participate in research assessments in English. Exclusion criteria were: had clearly experienced more than one discrete psychotic episode; diagnosis of bipolar disorder, psychotic depression, substance- induced psychotic disorder or current psychotic disorder due to a general medical condition; presence of current neurological disorders that would affect diagnosis or prognosis; clinically significant head trauma or other serious medical conditions that would significantly impair assessment, functioning or treatment.

3.c. What are the main exposure groups? In this article, each of the 404 participants can be considered the exposure group, as the data were collected upon study entrance. This article evaluated the choices made by community clinicians prior to the randomization that occurred with the NAVIGATE vs. community clinician choice group.

3.d. What is the main outcome (a.k.a., “primary endpoint”) of interest? How is it measured? Prescribing patterns and their adherence to recommended treatment principles in first episode patients, and how it relates to patients’ gender, racial background, ethnicity, age, cigarette smoking status, prior depression/anxiety symptoms, and the general location within the USA. Measures included antipsychotic prescription, prescription of more than one antipsychotic, long acting injectable antipsychotic prescription, first generation antipsychotic prescription, risperidone prescription and dose, olanzapine prescription and dose, and antidepressant prescription.

3.e. What additional efficacy/effectiveness outcomes (a.k.a., “secondary outcomes”) were assessed? In addition to evaluating the prescribing practices of community providers, the authors identified a subset of patients who may have benefited from changes to their psychotropic medication.

3.f. What is the main statistical approach? The data were analyzed using a “Bayesian perspective” for the correlate analyses, which do not require correction for multiple comparisons. In this method of analysis, “credible intervals” are determined, which are similar to confidence intervals. Instead of p values, Bayesian analyses provide a “posterior probability of being a risk factor” (PPRF). This method of analysis allows for multiple variables to be considered with a single choice, i.e. the gender, age, ethnicity, etc. of the patient on the final outcome, i.e. single atypical, addition of antidepressant, etc. This provides a measure of how strongly the evidence supports that a particular factor influenced the decided outcome.

3.g. What are the main results? The prescribing practices of community clinicians were varied in choice of medication, and in the influence of particular factors in treatment choice. Over 40 results of this analysis are included in the article. For brevity, the most notable results of prescribing patterns are as follows:


Secondary outcome measure results are as follows:


The most notable factors associated with prescribing patterns are as follows, where LE = lacking evidence; SE = some evidence; PE = positive evidence; +PE = strong evidence:



3.h. What did the authors conclude from their results? The majority of patients received an antipsychotic, suggesting that the need for antipsychotic medication is well recognized. Antipsychotic prescriptions were mostly concordant with guidelines, with the exception of olanzapine, which comprised 17% of the prescriptions. Due to the metabolic effects of olanzapine, PORT guidelines recommends against this agent in first-episode treatment. Additionally, olanzapine was more frequently prescribed at higher-than-recommended doses, and does not appear to be a result of treatment resistance. The authors hypothesize that the higher use of risperidone in younger patients may be a result of the FDA indication for adolescent treatment. Patients with insurance were more likely to be prescribed antipsychotics in concordance with first-episode guidelines. Diagnosis had little effect on prescription on treatment choice. Finally, smoking and substance use were not associated with risperidone or olanzapine dosing.


4.a. Was the study design appropriate to answer the research question? Yes, the data analyzed for this article were the result of patients treated by community providers prior to acceptance or knowledge of the study, which provides for a realistic sample. Additionally, the analysis using Bayesian statistics allowed for more direct probabilities, rather than the possibility of factors being erroneously connected to outcome.


4.b-c. Were the results internally/externally valid? Likely, yes, as the results were an analysis of decisions made prior to study entry. Limitations on this data include the fact that while multiple states were included in the study, it may not be generalizable as a true epidemiological sample. Additionally, the decisions made by the prescribing providers was not accessible to the authors, which may introduce the confounder that a patient was a “non-responder” to a different first choice medication, and the medication upon entry was not actually the community prescriber’s first choice treatment.


4.d. What are the major strengths/weakness of this report? The major strength of this report is that it utilizes data from a specific patient population (first episode psychosis patients) that were treated by community providers prior to the study, which provides for a real-life analysis of true prescribing practices not biased by knowledge of study inclusion. Advance knowledge of study inclusion may have resulted in a provider choosing a different treatment based on clinical guidelines. Realistic data allows for the greater psychiatric community to evaluate the community prescribing practices, and make positive change through provider education. As stated above, the major weakness of the paper is that the rationale behind provider choice was not reviewed. It is possible that in patients considered to have less than ideal regimens, the true first choice in treatment followed clinical guidelines; however, a failure of treatment resulted in a medication change that appeared to be the initial choice of the practitioner at the time of study enrollment.


5.a. What is your conclusion? This study gave a realistic snapshot of how community prescribers manage first episode psychosis patients. By taking data from patients without the advance knowledge of being enrolled, this study minimized provider’s bias of treatment choice. There are areas of concern about the prescribing practices, specifically the use of more than one antipsychotic, antidepressants without clear indication, and use of olanzapine at high doses—generally avoided in first-episode patients. These results are believable, and can be seen even in our patient population at Vanderbilt. However, it is also important to note that dealing with first episode patients can be particularly challenging, even for the seasoned provider, and despite some highlighted mismanagement, the majority of patients were prescribed an antipsychotic, an appropriate treatment choice. The factors affecting choice of medication were somewhat surprising in that patients with insurance were less likely to receive more than one antipsychotic. The authors briefly discuss the potential health policy implications of this outcome, however, this disparity should be discussed an evaluated in further detail, to find why insured patients receive care that is more adherent to clinical guidelines. One possible explanation is that there is more oversight in prescribing by insurance providers, in an effort to minimize cost and adhere to evidence-based medicine.


5.b. How could the results affect your practice? These results highlight the mismanagement of first-episode patients, and the different factors affecting community clinician choice. By recognizing the commonly made prescribing choices, and commonly made management errors, the inpatient physician can work towards more communication with outpatient providers about why the treatment was made for a particular patient, and what the hospital course revealed about the patient’s response to medication. This could help to reduce prescribing choices that are not within suggested management guidelines. Additionally, the inclusion of evidence-based practices in discharge summaries may help to educate community providers that are not as familiar with clinical guidelines.


5.c. How could the results affect health policy? The authors outline that the article has implications for health policy, specifically in regards to the potential for metabolic side effects of second generation antipsychotics, which appear to be utilized in this patient population. Stricter monitoring guidelines may be required to limit the adverse effects of these medications. Additionally, the disparity between insured and uninsured patient management is a target for policy change. The authors consider that significant changes in the health care delivery system, as well as changes in reimbursement involving evidence-based practice, may be necessary to improve provider practices with clinical guidelines.





Additional Resources


Section 1: Commonly Occurring Terms and Concepts


Bayesian Inference: Method of statistical analysis that does not require correction for multiple comparisons.


Credibility interval (CrI): Comparable to confidence interval


CSC: coordinated specialty care is a recovery-oriented treatment program for people with first episode psychosis. CSC uses a team of specialists who work with the client to create a personal treatment plan. The specialists offer psychotherapy, medication management geared to individuals with FEP, family education and support, case management, and work or education support, depending on the individual’s needs and preferences. CSC programs have been widely implemented in other countries, but not the US


ETP: Early treatment program, first study under RAISE initiative. “RAISE-ETP is a US nationwide effectiveness study conducted at 34 community treatment centers in 21 states for patients with first episode schizophrenia-spectrum disorders. Subjects had been treated with antipsychotics for 6 months or less at study entry


FEP: First Episode Psychosis


NAVIGATE: a coordinated specialty care treatment program for first episode psychosis that includes medical management guided by a decision support system, individual therapy, family psychoeducation, and supported employment and education, and community care, treatment determined by clinician choice


Posterior probability of being a risk factor (PPRF) (sometimes referred to as selected percentage): The probability that a variable is associated with an outcome. The larger the PPRF, the stronger the evidence for an association. For example, <50% is lacking evidence, 50–75% has some evidence, 75–95% positive evidence, 95–99% strong evidence and >99% very strong evidence


RAISE: Recovery After an Initial Schizophrenia Episode: a large-scale NIMH research initiative that began in 2008 with two studies examining different aspects of coordinated specialty care (CSC) treatments for people who were experiencing first episode psychosis. RAISE-ETP focused on whether or not the CSC treatment worked. The second, the RAISE Implementation and Evaluation Study (RAISE-IES), examined the best way for clinics to start using the CSC treatment program


Section 2: Key Points of Selected References


1) Robinson DG, Woerner MG, Delman HC, Kane JM. Pharmacological treatments for first-episode schizophrenia. Schizophr Bull. 2005; 31(3):705. [PubMed: 16006592]


2) Moore TA, Buchanan RW, Buckley PF, Chiles JA, Conley RR, Crismon ML, Essock SM, Finnerty M, Marder SR, Miller DD, McEvoy JP, Robinson DG, Schooler NS, Shon SP, Stroup TS, Miller AL. The Texas Medication Algorithm Project antipsychotic algorithm for schizophrenia: 2006 update. J Clin Psychiatry. 2007; 68(11):1751–62. [PubMed: 18052569]


3) Buchanan RW, Kreyenbuhl J, Kelly DL, Noel JM, Boggs DL, Fischer BA, Himelhoch S, Fang B, Peterson E, Aquino PR, Keller W. The 2009 schizophrenia PORT psychopharmacological treatment recommendations and summary statements. Schizophr Bull. 2010; 36(1):71–93. [PubMed: 19955390]


4) Barnes TRE. Evidence-based guidelines for the pharmacological treatment of schizophrenia: recommendations from the British Association for Psychopharmacology. J Psychopharmacol Oxf Engl. 2011 May; 25(5):567–620.