Integrity of Literature on Expressed Emotion and Relapse in Patients with Schizophrenia Verified by a p‐Curve Analysis

• Author: Julia Y. Carbonella,Daniel L. Hall,Amy Weisman de Mamani,Marc J. Weintraub,Jill M. Hooley
• DOI: http://doi.org/10.1111/famp.12208
• Date: 01 June 2017
• Published date: 01 June 2017

Integrity of Literature on Expressed Emotion and

Relapse in Patients with Schizophrenia Verified

by a p-Curve Analysis

MARC J. WEINTRAUB*

DANIEL L. HALL*

,†

JULIA Y. CARBONELLA*

AMY WEISMAN DE MAMANI*

JILL M. HOOLEY

‡

There is growing concern that much published research may have questionable val idity

due to phenomena such as publication bias and p-hacking. Within the psychiatric litera-

ture, the construct of expressed emotion (EE) is widely assumed to be a reliable predictor of

relapse across a range of mental illnesses. EE is an index of the family climate, measuring

how critical, hostile, and overinvolved a family member is toward a mental ly ill patient.

No study to date has examined the evidential value of this body of research as a whole.

That is to say, although many studies have shown a link between EE and symptom relapse,

the integrity of the literature from which this claim is derived has not been tested. In an

effort to confirm the integrity of the literature of EE predicting psychiatric relapse in

patients with schizophrenia, we conducted a p-curve analysis on all known studies examin-

ing EE (using the Camberwell Family Interview) to predict psychiatric relapse over a 9- to

12-month follow-up period. Results suggest that the body of literature on EE is unbiased

and has integrity, as there was a significant right skew of p-values, a nonsignificant left

skew of p-values, and a nonsignificant test of flatness. We conclude that EE is a robust and

valuable predictor of symptom relapse in schizophrenia.

Keywords: Expressed Emotion; Psychiatric Relapse; Schizophrenia; p-Curve Analysis

Fam Proc 56:436–444, 2017

INTRODUCTION

Expressed Emotion (EE) is a construct that reflects potentially problematic aspec ts of

the family environment, particularly for people with psychiatric disorders (Hooley,

2007). EE is a measure of how critical, hostile, or overly involved a family member is

toward an identified patient, and is thought to represent relational disturbances and

transactional patterns between the family member and that patient. EE is considered to

be a psychosocial stressor that interacts with patients’ diatheses, eventually culminating

in relapse (Hooley & Gotlib, 2000). The traditional method of measuring EE dichotomizes

family members into either high or low EE groupings based on the degree to which the

*Department of Psychology, University of Miami, Coral Gables, FL.

†

Department of Psychiatry, Massachusetts General Hospital/Harvard Medical School, Boston, MA.

‡

Department of Psychology, Harvard University, Cambridge, MA.

Correspondence concerning this article should be addressed to Marc J. Weintraub, Department of Psy-

chology, University of Miami, 5665 Ponce De Leon Blvd., Coral Gables, FL 33146. E-mail: mwein-

traub@psy.miami.edu

A special thanks to Michael McCullough, Ph.D., for his assistance on this manuscript.

436

Family Process, Vol. 56, No. 2, 2017 ©2016 Family Process Institute

doi: 10.1111/famp.12208