A team of researchers from the Department of Psychiatry and Behavioral Science at Stony Brook University has demonstrated that diagnoses of patients with psychotic symptoms shift significantly over time, raising concerns about the quality of patient treatment and research findings based solely on an initial diagnosis. The study, led by Distinguished Professor Evelyn Bromet, PhD was published in the June 15, 2011 American Journal of Psychiatry in Advance.
The findings are based on diagnostic assessments with 470 participants in the Suffolk County Mental Health project soon after their first admission for psychiatric hospital care, then at 6-month, 2-year and 10-year intervals. Even though the assessments were made by experienced psychiatrists using all of the information that accumulated over time, except the prior research diagnosis, fewer than half of the participants (49.3%) were given the same diagnosis at each of the four points in time. Participants with a diagnosis of schizophrenia or bipolar disorder were more likely to retain their initial diagnosis, while those with other diagnoses were more apt to be reclassified.
The authors note that some changes in diagnosis over time are to be expected. Diagnoses based on patterns of symptoms that unfold over time often cannot be made at the time of initial assessment, and sometimes there are gaps in the information available at the time of first admission. The fact that participants with a history of recent substance abuse were included in the study may have obscured the initial diagnosis.
However, these factors do not explain the magnitude of the shifts in diagnosis. The fact that half of the study population was misclassified at baseline using strict DSM criteria, is, the authors conclude, “a very concerning finding,” especially if the initial diagnosis is used as the sole basis for treatment or research.
One source of the problem, they point out, is that tests of the reliability of DSM criteria were based on inter-rater reliability at a single point in time, with little consideration of the reliability or predictive value of the criteria across time. The authors caution that, given the complexities involved, reassessing diagnoses over the long term is essential for clinical care and research.
In addition to Dr. Bromet, Roman Kotov, PhD, Laura Fochtmann, MD, Gabrielle Carlson, MD, and Marsha Tanenberg-Karant, MD from the Department of Psychiatry and Behavioral Science co-authored the article with Camilo Ruggero, PhD and Su-wei Chang, PhD. Many current and former members of the Department of Psychiatry were part of the diagnosis team over the years, including Drs. Eduardo Constantino, Thomas Craig, Frank Dowling, Shmuel Fennig, Silvana Fennig, Beatrice Kovasznay, Alan Miller, Bushra Naz, Joan Rubinstein, Carlos and Michele Pato, Ranganathan Ram, and Charles Rich. Janet Lavelle and Al Hamdy coordinated the study and conducted many interviews along with an able team of interviewers. According to Dr. Bromet, the initial decision to rediagnose participants over time grew out of conversations with Drs. Thomas Aronson, then in charge of the OPD, and Dr. Bruce Rosen at St. Catherine’s.